Release v3.9.16

Published from npm package build Source: https://github.com/thedotmack/claude-mem-source
Release v3.9.14
2025-10-06 22:55:16 -04:00 · 2025-10-03 21:47:35 -04:00 · 2025-10-03 21:45:36 -04:00 · 2025-10-03 21:09:11 -04:00 · 2025-10-03 20:55:36 -04:00 · 2025-10-03 18:27:36 -04:00
67 changed files with 5670 additions and 8005 deletions
@@ -1,5 +1,3 @@
-
-
                  GNU AFFERO GENERAL PUBLIC LICENSE
                      Version 3, 19 November 2007

@@ -1,86 +1,470 @@
-# 🧠 Claude Memory System (claude-mem)
+<div align="center">

-## Remember that one thing? Neither do we… but `claude-mem` does! 😵‍💫
+  <img src="claude-mem-logo-lm.webp#gh-light-mode-only" alt="claude-mem logo" width="360" height="auto" />
+  <img src="claude-mem-logo-dm.webp#gh-dark-mode-only" alt="claude-mem logo" width="360" height="auto" />

-Stop repeating yourself. `claude-mem` remembers what you and Claude Code figure out, so every new chat starts smarter than the last.
+  <p>
+    Memory compression and persistence system for Claude Code conversations
+  </p>

-## ⚡️ 10‑Second Setup

-```bash
-npm install -g claude-mem && claude-mem install
-```
+<!-- Badges -->
+<p>
+  <a href="https://www.npmjs.com/package/claude-mem">
+    <img src="https://img.shields.io/npm/v/claude-mem.svg" alt="npm version" />
+  </a>
+  <a href="https://github.com/thedotmack/claude-mem/graphs/contributors">
+    <img src="https://img.shields.io/github/contributors/thedotmack/claude-mem" alt="contributors" />
+  </a>
+  <a href="">
+    <img src="https://img.shields.io/github/last-commit/thedotmack/claude-mem" alt="last update" />
+  </a>
+  <a href="https://github.com/thedotmack/claude-mem/network/members">
+    <img src="https://img.shields.io/github/forks/thedotmack/claude-mem" alt="forks" />
+  </a>
+  <a href="https://github.com/thedotmack/claude-mem/stargazers">
+    <img src="https://img.shields.io/github/stars/thedotmack/claude-mem" alt="stars" />
+  </a>
+  <a href="https://github.com/thedotmack/claude-mem/issues/">
+    <img src="https://img.shields.io/github/issues/thedotmack/claude-mem" alt="open issues" />
+  </a>
+  <a href="https://github.com/thedotmack/claude-mem/blob/main/LICENSE">
+    <img src="https://img.shields.io/badge/license-AGPL--3.0-blue.svg" alt="license" />
+  </a>
+  <a href="https://nodejs.org/">
+    <img src="https://img.shields.io/badge/node-%3E%3D18.0.0-brightgreen.svg" alt="node version" />
+  </a>
+  <a href="https://modelcontextprotocol.io">
+    <img src="https://img.shields.io/badge/MCP-compatible-purple.svg" alt="MCP compatible" />
+  </a>
+  <a href="https://claude.com/claude-code">
+    <img src="https://img.shields.io/badge/Claude%20Code-enabled-orange.svg" alt="Claude Code enabled" />
+  </a>
+</p>

-That’s it. Restart Claude Code and you’re good. No config. No tedious setup or dependencies.
+<h4>
+    <a href="https://github.com/thedotmack/claude-mem">Documentation</a>
+  <span> · </span>
+    <a href="https://github.com/thedotmack/claude-mem/issues/">Report Bug</a>
+  <span> · </span>
+    <a href="https://github.com/thedotmack/claude-mem/issues/">Request Feature</a>
+  </h4>
+</div>

-## ✨ What You Get
+<br />

- Remembers key insights from your chats with Claude Code
- Starts new sessions with the right context
- Works quietly in the background
- One-command install and status check
+<!-- Table of Contents -->
+# :notebook_with_decorative_cover: Table of Contents

-## 🗑️ Smart Trash™ (Your Panic Button)
+- [About the Project](#star2-about-the-project)
+  * [Tech Stack](#space_invader-tech-stack)
+  * [Features](#dart-features)
+- [Getting Started](#toolbox-getting-started)
+  * [Prerequisites](#bangbang-prerequisites)
+  * [Installation](#gear-installation)
+  * [Running Tests](#test_tube-running-tests)
+- [Usage](#eyes-usage)
+  * [Basic Commands](#basic-commands)
+  * [Hook System](#hook-system)
+  * [Memory Operations](#memory-operations)
+  * [ChromaDB MCP Tools](#chromadb-mcp-tools)
+  * [Advanced Usage](#advanced-usage)
+- [Architecture](#building_construction-architecture)
+- [Configuration](#wrench-configuration)
+- [Roadmap](#compass-roadmap)
+- [Contributing](#wave-contributing)
+- [License](#warning-license)
+- [Contact](#handshake-contact)
+- [Acknowledgements](#gem-acknowledgements)

-Delete something by accident? It’s not gone.
- Everything goes to `~/.claude-mem/trash/`
- Restore with a single command: `claude-mem restore`
- Timestamped so you can see when things moved

-## 🎯 Why It’s Useful

- No more re-explaining your project over and over
- Pick up exactly where you left off
- Find past solutions fast when you face a familiar bug
- Your knowledge compounds the more you use it
+<!-- About the Project -->
+## :star2: About the Project

-## 🧭 Minimal Commands You’ll Ever Need
+claude-mem automatically captures, compresses, and retrieves context across Claude Code sessions, enabling true long-term memory through semantic search and intelligent compression.

-```bash
-claude-mem install          # Set up/repair integration
-claude-mem status           # Check everything’s working
-claude-mem load-context     # Peek at what it remembers
-claude-mem logs             # If you’re curious
-claude-mem uninstall        # Remove hooks
+Perfect for developers who want their AI assistant to remember project context, past decisions, and conversation history across sessions without manual context management.

-# Extras
-claude-mem trash-view       # See what’s in Smart Trash™
-claude-mem restore          # Restore deleted items
-```
+<!-- TechStack -->
+### :space_invader: Tech Stack

-## 📁 Where Stuff Lives (super simple)
+<details>
+  <summary>Core Technologies</summary>
+  <ul>
+    <li><a href="https://www.typescriptlang.org/">TypeScript</a></li>
+    <li><a href="https://nodejs.org/">Node.js</a></li>
+    <li><a href="https://bun.sh/">Bun</a></li>
+  </ul>
+</details>

-```
-~/.claude-mem/
-├── index/      # memory index
-├── archives/   # transcripts
-├── hooks/      # integration bits
-├── trash/      # Smart Trash™
-└── logs/       # diagnostics
-```
+<details>
+  <summary>Storage & Memory</summary>
+  <ul>
+    <li><a href="https://www.trychroma.com/">ChromaDB</a> - Vector database for semantic search</li>
+    <li><a href="https://www.sqlite.org/">SQLite</a> - Metadata and session tracking</li>
+    <li><a href="https://github.com/WiseLibs/better-sqlite3">better-sqlite3</a> - Fast SQLite bindings</li>
+  </ul>
+</details>

-## ✅ Requirements
+<details>
+<summary>AI & Integration</summary>
+  <ul>
+    <li><a href="https://github.com/anthropics/anthropic-sdk-typescript">Anthropic Agent SDK</a> - Async compression</li>
+    <li><a href="https://modelcontextprotocol.io">Model Context Protocol (MCP)</a> - Tool integration</li>
+    <li><a href="https://claude.com/claude-code">Claude Code</a> - Streaming hooks</li>
+  </ul>
+</details>

- Node.js 18+
- Claude Code
+<!-- Features -->
+### :dart: Features

-## 🆘 If Something’s Weird
+- :brain: **Automatic Memory Compression** - Real-time conversation capture and intelligent summarization
+- :mag: **Semantic Search** - ChromaDB-powered vector search for intelligent context retrieval
+- :package: **Project Isolation** - Memories segregated by project with multi-project support
+- :arrows_counterclockwise: **Session Persistence** - Context loads automatically at session start and `/clear` command
+- :dart: **MCP Integration** - 15+ ChromaDB tools via Model Context Protocol
+- :floppy_disk: **SQLite Storage** - Fast metadata and session tracking with embedded database
+- :wastebasket: **Smart Trash** - Safe file deletion with recovery capabilities
+- :zap: **Streaming Hooks** - Sub-50ms overhead for real-time event capture
+- :robot: **Agent SDK Compression** - Async transcript processing without blocking conversations
+- :bar_chart: **Session Overviews** - Automatic session summaries with temporal context

-```bash
-claude-mem status           # quick health check
-claude-mem install --force  # fixes most issues
-```
+<!-- Getting Started -->
+## :toolbox: Getting Started

-## 📄 License
+<!-- Prerequisites -->
+### :bangbang: Prerequisites

-Licensed under AGPL-3.0. See `LICENSE`.
+This project requires Node.js and works best with Claude Code

---
+- Node.js >= 18.0.0
+- Claude Code with MCP support
+- macOS/Linux (POSIX-compliant system)
+- Bun >= 1.0.0 (optional, for development)

-## Ready to remember more and repeat less?
+<!-- Installation -->
+### :gear: Installation
+
+Install claude-mem globally via npm

 ```bash
 npm install -g claude-mem
 claude-mem install
 ```

-Your future self will thank you. 🧠✨
+The interactive installer will guide you through three installation scopes:
+
+- **User** - Install for current user (default, recommended)
+- **Project** - Install for current project only
+- **Local** - Install to custom directory
+
+<!-- Running Tests -->
+### :test_tube: Running Tests
+
+To run tests, use the following commands
+
+```bash
+# Run all tests
+bun test
+
+# Run integration tests
+npm run test:integration
+
+# Run with coverage
+npm run test:coverage
+```
+
+<!-- Usage -->
+## :eyes: Usage
+
+### Basic Commands
+
+```bash
+# Check installation status
+claude-mem status
+
+# View operation logs
+claude-mem logs
+
+# Load context for current project
+claude-mem load-context --project my-project
+
+# View compressed memories (interactive)
+claude-mem restore
+
+# Manage trash bin
+claude-mem trash view
+claude-mem restore
+claude-mem trash empty
+```
+
+### Hook System
+
+claude-mem integrates with Claude Code via streaming hooks that capture conversation events:
+
+- **user-prompt-submit** - Captures user prompts in real-time
+- **post-tool-use** - Spawns Agent SDK for async compression
+- **stop-streaming** - Generates session overview and cleanup
+- **session-start** - Loads relevant context automatically
+
+Hooks are configured during installation with a 180-second timeout and run transparently in the background.
+
+### Memory Operations
+
+#### Manual Compression
+
+```bash
+claude-mem compress
+```
+
+Compress Claude Code transcripts into searchable memories with semantic embeddings.
+
+#### Context Loading
+
+```bash
+# Load last 10 memories for current project
+claude-mem load-context
+
+# Load specific number of memories
+claude-mem load-context --count 20
+
+# Filter by project
+claude-mem load-context --project my-app
+
+# Output raw JSON
+claude-mem load-context --raw
+```
+
+#### Trash Management
+
+claude-mem includes Smart Trash for safe file operations:
+
+```bash
+# Move files to trash
+claude-mem trash file.txt
+claude-mem trash -r directory/
+
+# View trash contents
+claude-mem trash view
+
+# Restore files interactively
+claude-mem restore
+
+# Empty trash permanently
+claude-mem trash empty
+```
+
+### ChromaDB MCP Tools
+
+claude-mem exposes 15+ ChromaDB operations via MCP:
+
+```bash
+# List collections
+claude-mem chroma-list-collections
+
+# Create collection
+claude-mem chroma-create-collection --collection-name memories
+
+# Query documents semantically
+claude-mem chroma-query-documents \
+  --collection-name memories \
+  --query-texts '["authentication implementation"]' \
+  --n-results 5
+
+# Add documents
+claude-mem chroma-add-documents \
+  --collection-name memories \
+  --documents '["content here"]' \
+  --ids '["mem-001"]'
+
+# Get documents by ID
+claude-mem chroma-get-documents \
+  --collection-name memories \
+  --ids '["mem-001"]'
+
+# Update documents
+claude-mem chroma-update-documents \
+  --collection-name memories \
+  --ids '["mem-001"]' \
+  --documents '["updated content"]'
+
+# Delete documents
+claude-mem chroma-delete-documents \
+  --collection-name memories \
+  --ids '["mem-001"]'
+```
+
+See all available Chroma MCP commands with `claude-mem --help`.
+
+### Advanced Usage
+
+#### Session Title Generation
+
+```bash
+# Generate title and subtitle from prompt
+claude-mem generate-title "implemented authentication with OAuth"
+
+# Output as JSON
+claude-mem generate-title "fixed bug in checkout" --json
+
+# Save to database
+claude-mem generate-title "added feature" --session-id abc123 --save
+```
+
+#### Diagnostics
+
+```bash
+# Run environment diagnostics
+claude-mem doctor
+
+# Output as JSON
+claude-mem doctor --json
+```
+
+#### Changelog Generation
+
+```bash
+# Generate changelog from memories
+claude-mem changelog
+
+# Preview without saving
+claude-mem changelog --preview
+
+# Generate for specific version
+claude-mem changelog --generate 3.9.0
+
+# Search historical versions
+claude-mem changelog --historical 5
+```
+
+## :building_construction: Architecture
+
+### Storage Structure
+
+```
+~/.claude-mem/
+├── archives/           # Compressed transcript backups
+├── chroma/            # ChromaDB vector database
+├── trash/             # Smart Trash with recovery
+├── hooks/             # Hook configurations
+├── logs/              # Operation logs
+└── claude-mem.db      # SQLite metadata database
+```
+
+### Memory System
+
+**Rolling Memory** - Real-time conversation turn capture via hooks with immediate ChromaDB storage
+
+**TranscriptCompressor** - Intelligent chunking and compression of large conversations
+
+**MCP Server** - 15+ ChromaDB tools for memory operations and semantic search
+
+**SQLite Backend** - Session tracking, metadata management, and diagnostics storage
+
+### Hook Integration
+
+Hooks communicate via JSON stdin/stdout and run with minimal overhead:
+
+1. **user-prompt-submit** - Stores user prompt immediately in ChromaDB
+2. **post-tool-use** - Spawns Agent SDK subprocess for async compression
+3. **stop-streaming** - Generates session overview, deletes SDK transcript
+4. **session-start** - Loads project-specific context invisibly
+
+### Project Structure
+
+```
+src/
+├── bin/           # CLI entry point
+├── commands/      # Command implementations
+├── core/          # Core compression logic
+├── services/      # SQLite, ChromaDB, path discovery
+├── shared/        # Configuration and utilities
+└── mcp-server.ts  # MCP server implementation
+
+hook-templates/    # Hook source files
+dist/              # Minified production bundle
+test/              # Unit and integration tests
+```
+
+## :wrench: Configuration
+
+### Hook Timeout
+
+Default hook timeout is 180 seconds. Configure during installation:
+
+```bash
+claude-mem install --timeout 300000  # 5 minutes
+```
+
+### MCP Server
+
+Skip MCP server installation if needed:
+
+```bash
+claude-mem install --skip-mcp
+```
+
+### Force Reinstall
+
+```bash
+claude-mem install --force
+```
+
+<!-- Roadmap -->
+## :compass: Roadmap
+
+* [x] Real-time conversation capture with streaming hooks
+* [x] ChromaDB vector storage for semantic search
+* [x] SQLite metadata and session tracking
+* [x] MCP server with 15+ ChromaDB tools
+* [x] Smart Trash for safe file deletion
+* [x] Automatic session overviews
+* [ ] Web UI for memory visualization
+* [ ] Cross-platform Windows support
+* [ ] Memory analytics and insights
+
+<!-- Contributing -->
+## :wave: Contributing
+
+<a href="https://github.com/thedotmack/claude-mem/graphs/contributors">
+  <img src="https://contrib.rocks/image?repo=thedotmack/claude-mem" />
+</a>
+
+Contributions are always welcome!
+
+1. Fork the repository
+2. Create a feature branch (`git checkout -b feature/AmazingFeature`)
+3. Commit your changes (`git commit -m 'Add some AmazingFeature'`)
+4. Push to the branch (`git push origin feature/AmazingFeature`)
+5. Open a Pull Request
+
+<!-- License -->
+## :warning: License
+
+Distributed under the AGPL-3.0 License. See [LICENSE](LICENSE) for more information.
+
+<!-- Contact -->
+## :handshake: Contact
+
+Alex Newman - [@thedotmack](https://github.com/thedotmack)
+
+Project Link: [https://github.com/thedotmack/claude-mem](https://github.com/thedotmack/claude-mem)
+
+NPM Package: [https://www.npmjs.com/package/claude-mem](https://www.npmjs.com/package/claude-mem)
+
+<!-- Acknowledgments -->
+## :gem: Acknowledgements
+
+ - [ChromaDB](https://www.trychroma.com/) - Vector database for AI applications
+ - [Anthropic](https://www.anthropic.com/) - Claude AI and Agent SDK
+ - [Model Context Protocol](https://modelcontextprotocol.io) - Standardized AI tool integration
+ - [better-sqlite3](https://github.com/WiseLibs/better-sqlite3) - Fast SQLite bindings
+ - [Shields.io](https://shields.io/) - Beautiful README badges
+ - [Awesome README Template](https://github.com/Louis3797/awesome-readme-template) - Template inspiration
+
+---
+
+**Philosophy**: claude-mem follows the **Make It Work First** approach - direct execution over defensive validation, natural failures instead of artificial guards, and memory as a living, evolving system. Context improves with use through semantic search, project isolation, and temporal relevance.
+
+**Built with TypeScript, ChromaDB, SQLite, and the Anthropic Agent SDK**
@@ -0,0 +1,265 @@
+# Windows Installation Guide
+
+## Overview
+
+Claude-mem now includes **experimental Windows support** as of v3.10.0 (October 2025). The cross-platform architecture uses the [Platform utility](/src/utils/platform.ts) for handling Windows-specific differences.
+
+## System Requirements
+
+- **Windows 10/11** (64-bit recommended)
+- **Node.js** >= 18.0.0
+- **PowerShell** 5.1 or later (for hook execution)
+- **Claude Code** with MCP support
+- **npm** package manager
+
+**Note**: Bun (>=1.0.0) is only required for development work on claude-mem itself, not for running it.
+
+## Known Windows-Specific Differences
+
+### 1. Hook Execution
+- **Unix/macOS**: Hooks execute via `/bin/sh`
+- **Windows**: Hooks execute via `powershell`
+
+All hooks are `.js` files that work on both platforms through Node.js.
+
+### 2. Path Handling
+Windows paths use backslashes (`\`) but Node.js normalizes these automatically. The data directory is located at:
+```
+C:\Users\YourUsername\.claude-mem\
+```
+
+### 3. File Permissions
+- **Unix/macOS**: Hook files get `chmod 755` permissions
+- **Windows**: Permission setting is a no-op (not needed)
+
+### 4. Smart Trash™ Alias
+The Smart Trash feature creates shell aliases differently:
+
+**Windows (PowerShell)**:
+```powershell
+# claude-mem smart trash alias
+function rm { claude-mem trash $args }
+```
+
+PowerShell profiles are located at:
+- `Documents\PowerShell\Microsoft.PowerShell_profile.ps1`
+- `Documents\WindowsPowerShell\Microsoft.PowerShell_profile.ps1`
+
+**Unix/macOS (Bash/Zsh)**:
+```bash
+# claude-mem smart trash alias
+alias rm='claude-mem trash'
+```
+
+### 5. UV Package Manager Installation
+The installer automatically installs `uv` using platform-specific methods:
+
+**Windows**:
+```powershell
+powershell -Command "irm https://astral.sh/uv/install.ps1 | iex"
+```
+
+**Unix/macOS**:
+```bash
+curl -LsSf https://astral.sh/uv/install.sh | sh
+```
+
+## Installation Steps
+
+### 1. Install claude-mem globally
+
+```powershell
+npm install -g claude-mem
+```
+
+### 2. Run the installer
+
+```powershell
+claude-mem install
+```
+
+The installer will:
+1. ✅ Create directory structure at `C:\Users\YourUsername\.claude-mem\`
+2. ✅ Install UV package manager via PowerShell
+3. ✅ Install Chroma MCP server for vector database
+4. ✅ Add CLAUDE.md instructions to `C:\Users\YourUsername\.claude\`
+5. ✅ Install slash commands (save.md, remember.md, claude-mem.md)
+6. ✅ Install memory hooks with PowerShell execution support
+7. ✅ Configure Claude Code settings
+
+### 3. Restart Claude Code
+
+After installation, restart Claude Code to activate the memory system.
+
+## Windows-Specific Caveats
+
+### PowerShell Execution Policy
+If hooks fail to execute, you may need to adjust your PowerShell execution policy:
+
+```powershell
+# Check current policy
+Get-ExecutionPolicy
+
+# Allow local scripts (choose one):
+Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser
+# OR for more security:
+Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope Process
+```
+
+### Path Environment Variable
+After installing UV, you may need to restart PowerShell or update your PATH:
+
+```powershell
+# Add to PATH if needed
+$env:PATH = "$env:USERPROFILE\.cargo\bin;$env:PATH"
+```
+
+The installer attempts to set this automatically, but it only affects the current process.
+
+### Better-SQLite3 Native Module
+The hooks use `better-sqlite3` for database access, which requires native compilation:
+- The installer runs `npm install` in the hooks directory
+- If this fails, hooks may still work if `better-sqlite3` is globally available
+- Build errors are silently ignored to prevent installation failure
+
+If you see database errors in logs:
+```powershell
+cd $env:USERPROFILE\.claude-mem\hooks
+npm install better-sqlite3 --build-from-source
+```
+
+## File Locations on Windows
+
+| Component | Windows Path |
+|-----------|-------------|
+| Data directory | `C:\Users\YourUsername\.claude-mem\` |
+| ChromaDB | `C:\Users\YourUsername\.claude-mem\chroma\` |
+| SQLite database | `C:\Users\YourUsername\.claude-mem\claude-mem.db` |
+| Hooks | `C:\Users\YourUsername\.claude-mem\hooks\` |
+| Archives | `C:\Users\YourUsername\.claude-mem\archives\` |
+| Trash | `C:\Users\YourUsername\.claude-mem\trash\` |
+| CLAUDE.md | `C:\Users\YourUsername\.claude\CLAUDE.md` |
+| Settings | `C:\Users\YourUsername\AppData\Roaming\Claude\settings.json` |
+| Commands | `C:\Users\YourUsername\.claude\commands\` |
+
+## Testing Your Installation
+
+### 1. Check installation status
+```powershell
+claude-mem status
+```
+
+### 2. Run diagnostics
+```powershell
+claude-mem doctor
+```
+
+This will verify:
+- ✅ Directory structure
+- ✅ Hook configuration
+- ✅ Database accessibility
+- ✅ MCP server integration
+
+### 3. Test memory storage
+Start Claude Code and have a brief conversation. Then check:
+```powershell
+claude-mem status
+```
+
+You should see memory counts increase.
+
+### 4. Search memories
+In Claude Code, ask:
+```
+Search my memories for [your topic]
+```
+
+Claude should use the `mcp__claude-mem__chroma_query_documents` tool automatically.
+
+## Known Issues & Limitations
+
+### ⚠️ PowerShell Profile Creation
+- **Issue**: PowerShell profiles may not exist by default
+- **Impact**: Smart Trash alias installation creates profile directories
+- **Status**: Installer handles this automatically
+
+### ⚠️ Native Module Compilation
+- **Issue**: `better-sqlite3` requires build tools
+- **Impact**: May need Visual Studio Build Tools installed
+- **Workaround**: Installer attempts silent installation; manually rebuild if needed
+
+### ⚠️ Path Case Sensitivity
+- **Issue**: Windows paths are case-insensitive but ChromaDB metadata stores exact case
+- **Impact**: Path comparisons may fail in edge cases
+- **Status**: Generally works fine; use consistent casing
+
+## Getting Help
+
+If you encounter Windows-specific issues:
+
+1. **Run diagnostics**:
+   ```powershell
+   claude-mem doctor
+   ```
+
+2. **Check logs**:
+   ```powershell
+   claude-mem logs
+   ```
+   Or view directly:
+   ```powershell
+   type $env:USERPROFILE\.claude-mem\logs\*.log
+   ```
+
+3. **Verify hook execution**:
+   Check Claude Code's hook output for errors during session start/stop
+
+4. **Report issues**:
+   - Include `claude-mem doctor` output
+   - Include relevant logs
+   - Specify Windows version and PowerShell version
+   - File at: https://github.com/thedotmack/claude-mem/issues
+
+## Development on Windows
+
+If you're developing claude-mem on Windows:
+
+### Prerequisites
+- Install Bun (Windows support is experimental): https://bun.sh/
+- Install Git for Windows
+- Install Visual Studio Build Tools for native modules
+
+### Build Commands
+```powershell
+# Install dependencies
+npm install
+
+# Build minified bundle
+npm run build
+
+# Link for local testing
+bun link
+
+# Reinstall with latest build
+claude-mem install --force
+```
+
+**Note**: Some npm scripts use Unix-style paths and may not work on Windows. Core development is recommended on Unix/macOS systems.
+
+## Summary
+
+✅ **Works on Windows**:
+- Memory capture via streaming hooks
+- ChromaDB semantic search
+- SQLite metadata storage
+- MCP server integration
+- All CLI commands
+- Smart Trash™ system
+- Automatic context loading
+
+⚠️ **May Require Extra Setup**:
+- PowerShell execution policy
+- Visual Studio Build Tools for native modules
+- Manual PATH configuration
+
+The core claude-mem functionality is fully operational on Windows, with the memory capture, storage, and retrieval systems working identically to Unix/macOS.
@@ -0,0 +1,150 @@
+#!/usr/bin/env node
+
+/**
+ * Post Tool Use Hook - Streaming SDK Version
+ *
+ * Feeds tool responses to the streaming SDK session for real-time processing.
+ * SDK decides what to store and calls bash commands directly.
+ */
+
+import path from 'path';
+import fs from 'fs';
+import { fileURLToPath } from 'url';
+import { query } from '@anthropic-ai/claude-agent-sdk';
+import { renderToolMessage, HOOK_CONFIG } from './shared/hook-prompt-renderer.js';
+import { getProjectName } from './shared/path-resolver.js';
+import { initializeDatabase, getActiveStreamingSessionsForProject } from './shared/hook-helpers.js';
+
+const __filename = fileURLToPath(import.meta.url);
+const __dirname = path.dirname(__filename);
+
+const HOOKS_LOG = path.join(process.env.HOME || '', '.claude-mem', 'logs', 'hooks.log');
+
+function debugLog(message, data = {}) {
+  if (process.env.CLAUDE_MEM_DEBUG === 'true') {
+    const timestamp = new Date().toISOString();
+    const logLine = `[${timestamp}] HOOK DEBUG: ${message} ${JSON.stringify(data)}\n`;
+    try {
+      fs.appendFileSync(HOOKS_LOG, logLine);
+      process.stderr.write(logLine);
+    } catch (error) {
+      // Silent fail on log errors
+    }
+  }
+}
+
+// Removed: buildStreamingToolMessage function
+// Now using centralized config from hook-prompt-renderer.js
+
+// =============================================================================
+// MAIN
+// =============================================================================
+
+let input = '';
+process.stdin.setEncoding('utf8');
+process.stdin.on('data', (chunk) => { input += chunk; });
+
+process.stdin.on('end', async () => {
+  let payload;
+  try {
+    payload = input ? JSON.parse(input) : {};
+  } catch (error) {
+    debugLog('PostToolUse: JSON parse error', { error: error.message });
+    console.log(JSON.stringify({ continue: true, suppressOutput: true }));
+    process.exit(0);
+  }
+
+  const { tool_name, tool_response, prompt, cwd, timestamp } = payload;
+  const project = cwd ? getProjectName(cwd) : 'unknown';
+
+  // Return immediately - process async in background (don't block next tool)
+  console.log(JSON.stringify({ async: true, asyncTimeout: 180000 }));
+
+  try {
+    // Load SDK session info from database
+    const db = initializeDatabase();
+
+    const sessions = getActiveStreamingSessionsForProject(db, project);
+    if (!sessions || sessions.length === 0) {
+      debugLog('PostToolUse: No streaming session found', { project });
+      db.close();
+      process.exit(0);
+    }
+
+    const sessionData = sessions[0];
+    const sdkSessionId = sessionData.sdk_session_id;
+
+    // Convert tool response to string
+    const toolResponseStr = typeof tool_response === 'string'
+      ? tool_response
+      : JSON.stringify(tool_response);
+
+    // Build message for SDK using centralized config
+    const message = renderToolMessage({
+      toolName: tool_name,
+      toolResponse: toolResponseStr,
+      userPrompt: prompt || '',
+      timestamp: timestamp || new Date().toISOString()
+    });
+
+    // Send to SDK and wait for processing to complete using centralized config
+    const response = query({
+      prompt: message,
+      options: {
+        model: HOOK_CONFIG.sdk.model,
+        resume: sdkSessionId,
+        allowedTools: HOOK_CONFIG.sdk.allowedTools,
+        maxTokens: HOOK_CONFIG.sdk.maxTokensTool,
+        cwd  // Must match where transcript was created
+      }
+    });
+
+    // Consume the stream to let SDK fully process
+    for await (const msg of response) {
+      debugLog('PostToolUse: SDK message', { type: msg.type, subtype: msg.subtype });
+
+      // SDK messages are structured differently than we expected
+      // - type: 'assistant' contains the assistant's response with content blocks
+      // - Content blocks can be text or tool_use
+      // - type: 'user' contains tool results
+      // - type: 'result' is the final summary
+
+      if (msg.type === 'assistant' && msg.message?.content) {
+        for (const block of msg.message.content) {
+          if (block.type === 'text') {
+            debugLog('PostToolUse: SDK text', { text: block.text?.slice(0, 200) });
+          } else if (block.type === 'tool_use') {
+            debugLog('PostToolUse: SDK tool_use', {
+              tool: block.name,
+              input: JSON.stringify(block.input).slice(0, 200)
+            });
+          }
+        }
+      } else if (msg.type === 'user' && msg.message?.content) {
+        for (const block of msg.message.content) {
+          if (block.type === 'tool_result') {
+            debugLog('PostToolUse: SDK tool_result', {
+              tool_use_id: block.tool_use_id,
+              content: typeof block.content === 'string' ? block.content.slice(0, 300) : JSON.stringify(block.content).slice(0, 300)
+            });
+          }
+        }
+      } else if (msg.type === 'result') {
+        debugLog('PostToolUse: SDK result', {
+          subtype: msg.subtype,
+          is_error: msg.is_error
+        });
+      }
+    }
+
+    debugLog('PostToolUse: SDK finished processing', { tool_name, sdkSessionId });
+
+    // Close database connection
+    db.close();
+  } catch (error) {
+    debugLog('PostToolUse: Error sending to SDK', { error: error.message });
+  }
+
+  // Exit cleanly after async processing completes
+  process.exit(0);
+});
@@ -0,0 +1,57 @@
+#!/usr/bin/env node
+
+/**
+ * Session Start Hook (SDK Version)
+ *
+ * Calls the CLI to load relevant context from ChromaDB at session start.
+ */
+
+import { createHookResponse, debugLog } from './shared/hook-helpers.js';
+
+// Read stdin
+let input = '';
+process.stdin.setEncoding('utf8');
+process.stdin.on('data', (chunk) => {
+  input += chunk;
+});
+
+process.stdin.on('end', async () => {
+  const payload = input ? JSON.parse(input) : {};
+
+  debugLog('SessionStart hook invoked (SDK version)', { cwd: payload.cwd });
+
+  const { cwd, source } = payload;
+
+  // Run on startup or /clear
+  if (source !== 'startup' && source !== 'clear') {
+    const response = createHookResponse('SessionStart', true);
+    console.log(JSON.stringify(response));
+    process.exit(0);
+  }
+
+  try {
+    // Call the CLI to load context
+    const { executeCliCommand } = await import('./shared/hook-helpers.js');
+
+    const result = await executeCliCommand('claude-mem', ['load-context', '--format', 'session-start']);
+
+    if (result.success && result.stdout) {
+      // Use the CLI output directly as context (it's already formatted)
+      const response = createHookResponse('SessionStart', true, {
+        context: result.stdout
+      });
+      console.log(JSON.stringify(response));
+      process.exit(0);
+    } else {
+      // Return without context
+      const response = createHookResponse('SessionStart', true);
+      console.log(JSON.stringify(response));
+      process.exit(0);
+    }
+  } catch (error) {
+    // Continue without context on error
+    const response = createHookResponse('SessionStart', true);
+    console.log(JSON.stringify(response));
+    process.exit(0);
+  }
+});
@@ -2,7 +2,7 @@

 /**
 * Hook Helper Functions
- * 
+ *
 * This module provides JavaScript wrappers around the TypeScript PromptOrchestrator
 * and HookTemplates system, making them accessible to the JavaScript hook scripts.
 */
@@ -10,6 +10,9 @@
 import { spawn } from 'child_process';
 import { join, dirname } from 'path';
 import { fileURLToPath } from 'url';
+import Database from 'better-sqlite3';
+import os from 'os';
+import fs from 'fs';

 const __dirname = dirname(fileURLToPath(import.meta.url));

@@ -39,30 +42,32 @@ export function createHookResponse(hookType, success, options = {}) {
  if (hookType === 'SessionStart') {
    if (success && options.context) {
      return {
-        continue: true,
-        suppressOutput: true,
        hookSpecificOutput: {
          hookEventName: 'SessionStart',
          additionalContext: options.context
        }
      };
-    } else if (success) {
-      // No context - just suppress output without any message
+    } else {
      return {
        continue: true,
        suppressOutput: true
      };
-    } else {
-      return {
-        continue: true, // Continue even on context loading failure
-        suppressOutput: true,
-        hookSpecificOutput: {
-          hookEventName: 'SessionStart',
-          additionalContext: `Context loading encountered an issue: ${options.error || 'Unknown error'}. Starting without previous context.`
-        }
-      };
    }
  }
+
+  if (hookType === 'UserPromptSubmit' || hookType === 'PostToolUse') {
+    return {
+      continue: true,
+      suppressOutput: true
+    };
+  }
+
+  if (hookType === 'Stop') {
+    return {
+      continue: true,
+      suppressOutput: true
+    };
+  }
  
  // Generic response for unknown hook types
  return {
@@ -115,9 +120,10 @@ export function formatSessionStartContext(contextData) {
 */
 export async function executeCliCommand(command, args = [], options = {}) {
  return new Promise((resolve) => {
+    const { input, ...spawnOptions } = options;
    const process = spawn(command, args, {
-      stdio: ['ignore', 'pipe', 'pipe'],
-      ...options
+      stdio: ['pipe', 'pipe', 'pipe'],
+      ...spawnOptions
    });
    
    let stdout = '';
@@ -135,6 +141,13 @@ export async function executeCliCommand(command, args = [], options = {}) {
      });
    }
    
+    if (input && process.stdin) {
+      process.stdin.write(input);
+      process.stdin.end();
+    } else if (process.stdin) {
+      process.stdin.end();
+    }
+
    process.on('close', (code) => {
      resolve({
        stdout: stdout.trim(),
@@ -224,4 +237,194 @@ export function debugLog(message, data = {}) {
    const timestamp = new Date().toISOString();
    console.error(`[${timestamp}] HOOK DEBUG: ${message}`, data);
  }
-}
+}
+
+// =============================================================================
+// DATABASE HELPERS (inline SQL to avoid 'claude-mem' import issues)
+// =============================================================================
+
+/**
+ * Get the claude-mem data directory path
+ */
+function getDataDirectory() {
+  return join(os.homedir(), '.claude-mem');
+}
+
+/**
+ * Get or create the database connection
+ */
+function getDatabase() {
+  const dataDir = getDataDirectory();
+  const dbPath = join(dataDir, 'claude-mem.db');
+
+  // Ensure directory exists
+  if (!fs.existsSync(dataDir)) {
+    fs.mkdirSync(dataDir, { recursive: true });
+  }
+
+  const db = new Database(dbPath);
+
+  // Apply optimized SQLite settings
+  db.pragma('journal_mode = WAL');
+  db.pragma('synchronous = NORMAL');
+  db.pragma('foreign_keys = ON');
+  db.pragma('temp_store = memory');
+
+  return db;
+}
+
+/**
+ * Ensure the streaming_sessions table exists
+ */
+function ensureStreamingSessionsTable(db) {
+  db.exec(`
+    CREATE TABLE IF NOT EXISTS streaming_sessions (
+      id INTEGER PRIMARY KEY AUTOINCREMENT,
+      claude_session_id TEXT UNIQUE NOT NULL,
+      sdk_session_id TEXT,
+      project TEXT NOT NULL,
+      title TEXT,
+      subtitle TEXT,
+      user_prompt TEXT,
+      started_at TEXT NOT NULL,
+      started_at_epoch INTEGER NOT NULL,
+      updated_at TEXT,
+      updated_at_epoch INTEGER,
+      completed_at TEXT,
+      completed_at_epoch INTEGER,
+      status TEXT NOT NULL CHECK(status IN ('active', 'completed', 'failed'))
+    )
+  `);
+
+  // Create indices if they don't exist
+  db.exec(`
+    CREATE INDEX IF NOT EXISTS idx_streaming_sessions_claude_id
+    ON streaming_sessions(claude_session_id)
+  `);
+  db.exec(`
+    CREATE INDEX IF NOT EXISTS idx_streaming_sessions_sdk_id
+    ON streaming_sessions(sdk_session_id)
+  `);
+  db.exec(`
+    CREATE INDEX IF NOT EXISTS idx_streaming_sessions_project_status
+    ON streaming_sessions(project, status)
+  `);
+}
+
+/**
+ * Create a new streaming session record
+ */
+export function createStreamingSession(db, { claude_session_id, project, user_prompt, started_at }) {
+  ensureStreamingSessionsTable(db);
+
+  const timestamp = started_at || new Date().toISOString();
+  const epoch = new Date(timestamp).getTime();
+
+  const stmt = db.prepare(`
+    INSERT INTO streaming_sessions (
+      claude_session_id, project, user_prompt, started_at, started_at_epoch, status
+    ) VALUES (?, ?, ?, ?, ?, 'active')
+  `);
+
+  const info = stmt.run(claude_session_id, project, user_prompt || null, timestamp, epoch);
+
+  return db.prepare('SELECT * FROM streaming_sessions WHERE id = ?').get(info.lastInsertRowid);
+}
+
+/**
+ * Update a streaming session by internal ID
+ */
+export function updateStreamingSession(db, id, updates) {
+  const timestamp = new Date().toISOString();
+  const epoch = Date.now();
+
+  const parts = [];
+  const values = [];
+
+  if (updates.sdk_session_id !== undefined) {
+    parts.push('sdk_session_id = ?');
+    values.push(updates.sdk_session_id);
+  }
+  if (updates.title !== undefined) {
+    parts.push('title = ?');
+    values.push(updates.title);
+  }
+  if (updates.subtitle !== undefined) {
+    parts.push('subtitle = ?');
+    values.push(updates.subtitle);
+  }
+  if (updates.status !== undefined) {
+    parts.push('status = ?');
+    values.push(updates.status);
+  }
+  if (updates.completed_at !== undefined) {
+    const completedTimestamp = typeof updates.completed_at === 'string'
+      ? updates.completed_at
+      : new Date(updates.completed_at).toISOString();
+    const completedEpoch = new Date(completedTimestamp).getTime();
+    parts.push('completed_at = ?', 'completed_at_epoch = ?');
+    values.push(completedTimestamp, completedEpoch);
+  }
+
+  // Always update the updated_at timestamp
+  parts.push('updated_at = ?', 'updated_at_epoch = ?');
+  values.push(timestamp, epoch);
+
+  values.push(id);
+
+  const stmt = db.prepare(`
+    UPDATE streaming_sessions
+    SET ${parts.join(', ')}
+    WHERE id = ?
+  `);
+
+  stmt.run(...values);
+
+  return db.prepare('SELECT * FROM streaming_sessions WHERE id = ?').get(id);
+}
+
+/**
+ * Get active streaming sessions for a project
+ */
+export function getActiveStreamingSessionsForProject(db, project) {
+  ensureStreamingSessionsTable(db);
+
+  const stmt = db.prepare(`
+    SELECT * FROM streaming_sessions
+    WHERE project = ? AND status = 'active'
+    ORDER BY started_at_epoch DESC
+  `);
+
+  return stmt.all(project);
+}
+
+/**
+ * Mark a session as completed
+ */
+export function markStreamingSessionCompleted(db, id) {
+  const timestamp = new Date().toISOString();
+  const epoch = Date.now();
+
+  const stmt = db.prepare(`
+    UPDATE streaming_sessions
+    SET status = ?,
+        completed_at = ?,
+        completed_at_epoch = ?,
+        updated_at = ?,
+        updated_at_epoch = ?
+    WHERE id = ?
+  `);
+
+  stmt.run('completed', timestamp, epoch, timestamp, epoch, id);
+
+  return db.prepare('SELECT * FROM streaming_sessions WHERE id = ?').get(id);
+}
+
+/**
+ * Initialize database with migrations and return connection
+ */
+export function initializeDatabase() {
+  const db = getDatabase();
+  ensureStreamingSessionsTable(db);
+  return db;
+}
@@ -0,0 +1,278 @@
+// src/prompts/hook-prompts.config.ts
+var HOOK_CONFIG = {
+  maxUserPromptLength: 200,
+  maxToolResponseLength: 20000,
+  sdk: {
+    model: "claude-sonnet-4-5",
+    allowedTools: ["Bash"],
+    maxTokensSystem: 8192,
+    maxTokensTool: 8192,
+    maxTokensEnd: 2048
+  }
+};
+var SYSTEM_PROMPT = `You are a semantic memory compressor for claude-mem. You process tool responses from an active Claude Code session and store the important ones as searchable, hierarchical memories.
+
+# SESSION CONTEXT
+- Project: {{project}}
+- Session: {{sessionId}}
+- Date: {{date}}
+- User Request: "{{userPrompt}}"
+
+# YOUR JOB
+
+## FIRST: Generate Session Title
+
+IMMEDIATELY generate a title and subtitle for this session based on the user request.
+
+Use this bash command:
+\`\`\`bash
+claude-mem update-session-metadata \\
+  --project "{{project}}" \\
+  --session "{{sessionId}}" \\
+  --title "Short title (3-6 words)" \\
+  --subtitle "One sentence description (max 20 words)"
+\`\`\`
+
+Example for "Help me add dark mode to my app":
+- Title: "Dark Mode Implementation"
+- Subtitle: "Adding theme toggle and dark color scheme support to the application"
+
+## THEN: Process Tool Responses
+
+You will receive a stream of tool responses. For each one:
+
+1. ANALYZE: Does this contain information worth remembering?
+2. DECIDE: Should I store this or skip it?
+3. EXTRACT: What are the key semantic concepts?
+4. DECOMPOSE: Break into title + subtitle + atomic facts + narrative
+5. STORE: Use bash to save the hierarchical memory
+6. TRACK: Keep count of stored memories (001, 002, 003...)
+
+# WHAT TO STORE
+
+Store these:
+- File contents with logic, algorithms, or patterns
+- Search results revealing project structure
+- Build errors or test failures with context
+- Code revealing architecture or design decisions
+- Git diffs with significant changes
+- Command outputs showing system state
+
+Skip these:
+- Simple status checks (git status with no changes)
+- Trivial edits (one-line config changes)
+- Repeated operations
+- Binary data or noise
+- Anything without semantic value
+
+# HIERARCHICAL MEMORY FORMAT
+
+Each memory has FOUR components:
+
+## 1. TITLE (3-8 words)
+A scannable headline that captures the core action or topic.
+Examples:
+- "SDK Transcript Cleanup Implementation"
+- "Hook System Architecture Analysis"
+- "ChromaDB Migration Planning"
+
+## 2. SUBTITLE (max 24 words)
+A concise, memorable summary that captures the essence of the change.
+Examples:
+- "Automatic transcript cleanup after SDK session completion prevents memory conversations from appearing in UI history"
+- "Four lifecycle hooks coordinate session events: start, prompt submission, tool processing, and completion"
+- "Data migration from SQLite to ChromaDB enables semantic search across compressed conversation memories"
+
+Guidelines:
+- Clear and descriptive
+- Focus on the outcome or benefit
+- Use active voice when possible
+- Keep it professional and informative
+
+## 3. ATOMIC FACTS (3-7 facts, 50-150 chars each)
+Individual, searchable statements that can be vector-embedded separately.
+Each fact is ONE specific piece of information.
+
+Examples:
+- "stop-streaming.js: Auto-deletes SDK transcripts after completion"
+- "Path format: ~/.claude/projects/{sanitized-cwd}/{sessionId}.jsonl"
+- "Uses fs.unlink with graceful error handling for missing files"
+- "Checks two transcript path formats for backward compatibility"
+
+Guidelines:
+- Start with filename or component when relevant
+- Be specific: include paths, function names, actual values
+- Each fact stands alone (no pronouns like "it" or "this")
+- 50-150 characters target
+- Focus on searchable technical details
+
+## 4. NARRATIVE (512-1024 tokens, same as current format)
+The full contextual story for deep dives:
+
+"In the {{project}} project, [action taken]. [Technical details: files, functions, concepts]. [Why this matters]."
+
+This is the detailed explanation for when someone needs full context.
+
+# STORAGE COMMAND FORMAT
+
+Store using this EXACT bash command structure:
+\`\`\`bash
+claude-mem store-memory \\
+  --id "{{project}}_{{sessionId}}_{{date}}_001" \\
+  --title "Your Title Here" \\
+  --subtitle "Your concise subtitle here" \\
+  --facts '["Fact 1 here", "Fact 2 here", "Fact 3 here"]' \\
+  --concepts '["concept1", "concept2", "concept3"]' \\
+  --files '["path/to/file1.js", "path/to/file2.ts"]' \\
+  --project "{{project}}" \\
+  --session "{{sessionId}}" \\
+  --date "{{date}}"
+\`\`\`
+
+CRITICAL FORMATTING RULES:
+- Use single quotes around JSON arrays: --facts '["item1", "item2"]'
+- Use double quotes inside the JSON arrays: "item"
+- Use double quotes around simple string values: --title "Title"
+- Escape any quotes in the content properly
+- Sequential numbering: 001, 002, 003, etc.
+
+Concepts: 2-5 broad categories (e.g., "hooks", "storage", "async-processing")
+Files: Actual file paths touched (e.g., "hooks/stop-streaming.js")
+
+# EXAMPLE MEMORY
+
+Tool response shows: [Read file hooks/stop-streaming.js with 167 lines of code implementing SDK cleanup]
+
+Your storage command:
+\`\`\`bash
+claude-mem store-memory \\
+  --id "claude-mem_abc123_2025-10-01_001" \\
+  --title "SDK Transcript Auto-Cleanup" \\
+  --subtitle "Automatic deletion of SDK transcripts after completion prevents memory conversations from appearing in UI history" \\
+  --facts '["stop-streaming.js: Deletes SDK transcript after overview generation", "Path: ~/.claude/projects/{sanitized-cwd}/{sessionId}.jsonl", "Uses fs.unlink with error handling for missing files", "Prevents memory conversations from polluting Claude Code UI"]' \\
+  --concepts '["cleanup", "SDK-lifecycle", "UX", "file-management"]' \\
+  --files '["hooks/stop-streaming.js"]' \\
+  --project "claude-mem" \\
+  --session "abc123" \\
+  --date "2025-10-01"
+\`\`\`
+
+# STATE TRACKING
+
+CRITICAL: Keep track of your memory counter across all tool messages.
+- Start at 001
+- Increment for each stored memory
+- Never repeat numbers
+- Each session has separate numbering
+
+# SESSION END
+
+At the end (when I send "SESSION ENDING"), generate an overview using:
+\`\`\`bash
+claude-mem store-overview --project "{{project}}" --session "{{sessionId}}" --content "2-3 sentence overview"
+\`\`\`
+
+# IMPORTANT REMINDERS
+
+- You're processing a DIFFERENT Claude Code session (not your own)
+- Use Bash tool to call claude-mem commands
+- Keep subtitles clear and informative (max 24 words)
+- Each fact is ONE specific thing (not multiple ideas)
+- Be selective - quality over quantity
+- Always increment memory numbers
+- Facts should be searchable (specific file names, paths, functions)
+
+Ready for tool responses.`;
+var TOOL_MESSAGE = `# Tool Response {{timeFormatted}}
+
+Tool: {{toolName}}
+User Context: "{{userPrompt}}"
+
+\`\`\`
+{{toolResponse}}
+\`\`\`
+
+Analyze and store if meaningful.`;
+var END_MESSAGE = `# SESSION ENDING
+
+Review our entire conversation. Generate a concise 2-3 sentence overview of what was accomplished.
+
+Store it using Bash:
+\`\`\`bash
+claude-mem store-overview --project "{{project}}" --session "{{sessionId}}" --content "YOUR_OVERVIEW_HERE"
+\`\`\`
+
+Focus on: what was done, current state, key decisions, outcomes.`;
+var PROMPTS = {
+  system: SYSTEM_PROMPT,
+  tool: TOOL_MESSAGE,
+  end: END_MESSAGE
+};
+
+// src/prompts/hook-prompt-renderer.ts
+function substituteVariables(template, variables) {
+  let result = template;
+  for (const [key, value] of Object.entries(variables)) {
+    const placeholder = `{{${key}}}`;
+    result = result.split(placeholder).join(value);
+  }
+  return result;
+}
+function truncate(text, maxLength) {
+  if (text.length <= maxLength)
+    return text;
+  return text.slice(0, maxLength) + (text.length > maxLength ? "..." : "");
+}
+function formatTime(timestamp) {
+  const timePart = timestamp.split("T")[1];
+  if (!timePart)
+    return "";
+  return timePart.slice(0, 8);
+}
+function renderSystemPrompt(variables) {
+  const userPromptTruncated = truncate(variables.userPrompt, HOOK_CONFIG.maxUserPromptLength);
+  return substituteVariables(PROMPTS.system, {
+    project: variables.project,
+    sessionId: variables.sessionId,
+    date: variables.date,
+    userPrompt: userPromptTruncated
+  });
+}
+function renderToolMessage(variables) {
+  const userPromptTruncated = truncate(variables.userPrompt, HOOK_CONFIG.maxUserPromptLength);
+  const toolResponseTruncated = truncate(variables.toolResponse, HOOK_CONFIG.maxToolResponseLength);
+  const timeFormatted = formatTime(variables.timestamp);
+  return substituteVariables(PROMPTS.tool, {
+    toolName: variables.toolName,
+    toolResponse: toolResponseTruncated,
+    userPrompt: userPromptTruncated,
+    timestamp: variables.timestamp,
+    timeFormatted
+  });
+}
+function renderEndMessage(variables) {
+  return substituteVariables(PROMPTS.end, {
+    project: variables.project,
+    sessionId: variables.sessionId
+  });
+}
+function renderPrompt(type, variables) {
+  switch (type) {
+    case "system":
+      return renderSystemPrompt(variables);
+    case "tool":
+      return renderToolMessage(variables);
+    case "end":
+      return renderEndMessage(variables);
+    default:
+      throw new Error(`Unknown prompt type: ${type}`);
+  }
+}
+export {
+  renderToolMessage,
+  renderSystemPrompt,
+  renderPrompt,
+  renderEndMessage,
+  PROMPTS,
+  HOOK_CONFIG
+};
@@ -0,0 +1,217 @@
+// src/prompts/hook-prompts.config.ts
+var HOOK_CONFIG = {
+  maxUserPromptLength: 200,
+  maxToolResponseLength: 20000,
+  sdk: {
+    model: "claude-sonnet-4-5",
+    allowedTools: ["Bash"],
+    maxTokensSystem: 8192,
+    maxTokensTool: 8192,
+    maxTokensEnd: 2048
+  }
+};
+var SYSTEM_PROMPT = `You are a semantic memory compressor for claude-mem. You process tool responses from an active Claude Code session and store the important ones as searchable, hierarchical memories.
+
+# SESSION CONTEXT
+- Project: {{project}}
+- Session: {{sessionId}}
+- Date: {{date}}
+- User Request: "{{userPrompt}}"
+
+# YOUR JOB
+
+## FIRST: Generate Session Title
+
+IMMEDIATELY generate a title and subtitle for this session based on the user request.
+
+Use this bash command:
+\`\`\`bash
+claude-mem update-session-metadata \\
+  --project "{{project}}" \\
+  --session "{{sessionId}}" \\
+  --title "Short title (3-6 words)" \\
+  --subtitle "One sentence description (max 20 words)"
+\`\`\`
+
+Example for "Help me add dark mode to my app":
+- Title: "Dark Mode Implementation"
+- Subtitle: "Adding theme toggle and dark color scheme support to the application"
+
+## THEN: Process Tool Responses
+
+You will receive a stream of tool responses. For each one:
+
+1. ANALYZE: Does this contain information worth remembering?
+2. DECIDE: Should I store this or skip it?
+3. EXTRACT: What are the key semantic concepts?
+4. DECOMPOSE: Break into title + subtitle + atomic facts + narrative
+5. STORE: Use bash to save the hierarchical memory
+6. TRACK: Keep count of stored memories (001, 002, 003...)
+
+# WHAT TO STORE
+
+Store these:
+- File contents with logic, algorithms, or patterns
+- Search results revealing project structure
+- Build errors or test failures with context
+- Code revealing architecture or design decisions
+- Git diffs with significant changes
+- Command outputs showing system state
+
+Skip these:
+- Simple status checks (git status with no changes)
+- Trivial edits (one-line config changes)
+- Repeated operations
+- Binary data or noise
+- Anything without semantic value
+
+# HIERARCHICAL MEMORY FORMAT
+
+Each memory has FOUR components:
+
+## 1. TITLE (3-8 words)
+A scannable headline that captures the core action or topic.
+Examples:
+- "SDK Transcript Cleanup Implementation"
+- "Hook System Architecture Analysis"
+- "ChromaDB Migration Planning"
+
+## 2. SUBTITLE (max 24 words)
+A concise, memorable summary that captures the essence of the change.
+Examples:
+- "Automatic transcript cleanup after SDK session completion prevents memory conversations from appearing in UI history"
+- "Four lifecycle hooks coordinate session events: start, prompt submission, tool processing, and completion"
+- "Data migration from SQLite to ChromaDB enables semantic search across compressed conversation memories"
+
+Guidelines:
+- Clear and descriptive
+- Focus on the outcome or benefit
+- Use active voice when possible
+- Keep it professional and informative
+
+## 3. ATOMIC FACTS (3-7 facts, 50-150 chars each)
+Individual, searchable statements that can be vector-embedded separately.
+Each fact is ONE specific piece of information.
+
+Examples:
+- "stop-streaming.js: Auto-deletes SDK transcripts after completion"
+- "Path format: ~/.claude/projects/{sanitized-cwd}/{sessionId}.jsonl"
+- "Uses fs.unlink with graceful error handling for missing files"
+- "Checks two transcript path formats for backward compatibility"
+
+Guidelines:
+- Start with filename or component when relevant
+- Be specific: include paths, function names, actual values
+- Each fact stands alone (no pronouns like "it" or "this")
+- 50-150 characters target
+- Focus on searchable technical details
+
+## 4. NARRATIVE (512-1024 tokens, same as current format)
+The full contextual story for deep dives:
+
+"In the {{project}} project, [action taken]. [Technical details: files, functions, concepts]. [Why this matters]."
+
+This is the detailed explanation for when someone needs full context.
+
+# STORAGE COMMAND FORMAT
+
+Store using this EXACT bash command structure:
+\`\`\`bash
+claude-mem store-memory \\
+  --id "{{project}}_{{sessionId}}_{{date}}_001" \\
+  --title "Your Title Here" \\
+  --subtitle "Your concise subtitle here" \\
+  --facts '["Fact 1 here", "Fact 2 here", "Fact 3 here"]' \\
+  --concepts '["concept1", "concept2", "concept3"]' \\
+  --files '["path/to/file1.js", "path/to/file2.ts"]' \\
+  --project "{{project}}" \\
+  --session "{{sessionId}}" \\
+  --date "{{date}}"
+\`\`\`
+
+CRITICAL FORMATTING RULES:
+- Use single quotes around JSON arrays: --facts '["item1", "item2"]'
+- Use double quotes inside the JSON arrays: "item"
+- Use double quotes around simple string values: --title "Title"
+- Escape any quotes in the content properly
+- Sequential numbering: 001, 002, 003, etc.
+
+Concepts: 2-5 broad categories (e.g., "hooks", "storage", "async-processing")
+Files: Actual file paths touched (e.g., "hooks/stop-streaming.js")
+
+# EXAMPLE MEMORY
+
+Tool response shows: [Read file hooks/stop-streaming.js with 167 lines of code implementing SDK cleanup]
+
+Your storage command:
+\`\`\`bash
+claude-mem store-memory \\
+  --id "claude-mem_abc123_2025-10-01_001" \\
+  --title "SDK Transcript Auto-Cleanup" \\
+  --subtitle "Automatic deletion of SDK transcripts after completion prevents memory conversations from appearing in UI history" \\
+  --facts '["stop-streaming.js: Deletes SDK transcript after overview generation", "Path: ~/.claude/projects/{sanitized-cwd}/{sessionId}.jsonl", "Uses fs.unlink with error handling for missing files", "Prevents memory conversations from polluting Claude Code UI"]' \\
+  --concepts '["cleanup", "SDK-lifecycle", "UX", "file-management"]' \\
+  --files '["hooks/stop-streaming.js"]' \\
+  --project "claude-mem" \\
+  --session "abc123" \\
+  --date "2025-10-01"
+\`\`\`
+
+# STATE TRACKING
+
+CRITICAL: Keep track of your memory counter across all tool messages.
+- Start at 001
+- Increment for each stored memory
+- Never repeat numbers
+- Each session has separate numbering
+
+# SESSION END
+
+At the end (when I send "SESSION ENDING"), generate an overview using:
+\`\`\`bash
+claude-mem store-overview --project "{{project}}" --session "{{sessionId}}" --content "2-3 sentence overview"
+\`\`\`
+
+# IMPORTANT REMINDERS
+
+- You're processing a DIFFERENT Claude Code session (not your own)
+- Use Bash tool to call claude-mem commands
+- Keep subtitles clear and informative (max 24 words)
+- Each fact is ONE specific thing (not multiple ideas)
+- Be selective - quality over quantity
+- Always increment memory numbers
+- Facts should be searchable (specific file names, paths, functions)
+
+Ready for tool responses.`;
+var TOOL_MESSAGE = `# Tool Response {{timeFormatted}}
+
+Tool: {{toolName}}
+User Context: "{{userPrompt}}"
+
+\`\`\`
+{{toolResponse}}
+\`\`\`
+
+Analyze and store if meaningful.`;
+var END_MESSAGE = `# SESSION ENDING
+
+Review our entire conversation. Generate a concise 2-3 sentence overview of what was accomplished.
+
+Store it using Bash:
+\`\`\`bash
+claude-mem store-overview --project "{{project}}" --session "{{sessionId}}" --content "YOUR_OVERVIEW_HERE"
+\`\`\`
+
+Focus on: what was done, current state, key decisions, outcomes.`;
+var PROMPTS = {
+  system: SYSTEM_PROMPT,
+  tool: TOOL_MESSAGE,
+  end: END_MESSAGE
+};
+export {
+  TOOL_MESSAGE,
+  SYSTEM_PROMPT,
+  PROMPTS,
+  HOOK_CONFIG,
+  END_MESSAGE
+};
@@ -0,0 +1,108 @@
+#!/usr/bin/env node
+
+/**
+ * Path resolver utility for Claude Memory hooks
+ * Provides proper path handling using environment variables
+ */
+
+import { join, basename } from 'path';
+import { homedir } from 'os';
+
+/**
+ * Gets the base data directory for claude-mem
+ * @returns {string} Data directory path
+ */
+export function getDataDir() {
+  return process.env.CLAUDE_MEM_DATA_DIR || join(homedir(), '.claude-mem');
+}
+
+/**
+ * Gets the settings file path
+ * @returns {string} Settings file path
+ */
+export function getSettingsPath() {
+  return join(getDataDir(), 'settings.json');
+}
+
+/**
+ * Gets the archives directory path
+ * @returns {string} Archives directory path
+ */
+export function getArchivesDir() {
+  return process.env.CLAUDE_MEM_ARCHIVES_DIR || join(getDataDir(), 'archives');
+}
+
+/**
+ * Gets the logs directory path
+ * @returns {string} Logs directory path
+ */
+export function getLogsDir() {
+  return process.env.CLAUDE_MEM_LOGS_DIR || join(getDataDir(), 'logs');
+}
+
+/**
+ * Gets the compact flag file path
+ * @returns {string} Compact flag file path
+ */
+export function getCompactFlagPath() {
+  return join(getDataDir(), '.compact-running');
+}
+
+/**
+ * Gets the claude-mem package root directory
+ * @returns {Promise<string>} Package root path
+ */
+export async function getPackageRoot() {
+  // Method 1: Check if we're running from development
+  const devPath = join(homedir(), 'Scripts', 'claude-mem-source');
+  const { existsSync } = await import('fs');
+  if (existsSync(join(devPath, 'package.json'))) {
+    return devPath;
+  }
+
+  // Method 2: Follow the binary symlink
+  try {
+    const { execSync } = await import('child_process');
+    const { realpathSync } = await import('fs');
+    const binPath = execSync('which claude-mem', { encoding: 'utf8' }).trim();
+    const realBinPath = realpathSync(binPath);
+    // Binary is typically at package_root/dist/claude-mem.min.js
+    return join(realBinPath, '../..');
+  } catch {}
+
+  throw new Error('Cannot locate claude-mem package root');
+}
+
+/**
+ * Gets the project root directory
+ * Uses CLAUDE_PROJECT_DIR environment variable if available, otherwise falls back to cwd
+ * @returns {string} Project root path
+ */
+export function getProjectRoot() {
+  return process.env.CLAUDE_PROJECT_DIR || process.cwd();
+}
+
+/**
+ * Derives project name from CLAUDE_PROJECT_DIR or current working directory
+ * Priority: CLAUDE_PROJECT_DIR > cwd parameter > process.cwd()
+ * @param {string} [cwd] - Optional current working directory from hook payload
+ * @returns {string} Project name (basename of project directory)
+ */
+export function getProjectName(cwd) {
+  const projectRoot = process.env.CLAUDE_PROJECT_DIR || cwd || process.cwd();
+  return basename(projectRoot);
+}
+
+/**
+ * Gets all common paths used by hooks
+ * @returns {Object} Object containing all common paths
+ */
+export function getPaths() {
+  return {
+    dataDir: getDataDir(),
+    settingsPath: getSettingsPath(),
+    archivesDir: getArchivesDir(),
+    logsDir: getLogsDir(),
+    compactFlagPath: getCompactFlagPath()
+  };
+}
@@ -0,0 +1,134 @@
+#!/usr/bin/env node
+
+/**
+ * Stop Hook - Simple Orchestrator
+ *
+ * Signals session end to SDK, which generates and stores the overview via CLI.
+ * Cleans up SDK transcript from UI.
+ */
+
+import path from 'path';
+import fs from 'fs';
+import { query } from '@anthropic-ai/claude-agent-sdk';
+import { renderEndMessage, HOOK_CONFIG } from './shared/hook-prompt-renderer.js';
+import { getProjectName } from './shared/path-resolver.js';
+import { initializeDatabase, getActiveStreamingSessionsForProject, markStreamingSessionCompleted } from './shared/hook-helpers.js';
+
+const HOOKS_LOG = path.join(process.env.HOME || '', '.claude-mem', 'logs', 'hooks.log');
+
+function debugLog(message, data = {}) {
+  if (process.env.CLAUDE_MEM_DEBUG === 'true') {
+    const timestamp = new Date().toISOString();
+    const logLine = `[${timestamp}] HOOK DEBUG: ${message} ${JSON.stringify(data)}\n`;
+    try {
+      fs.appendFileSync(HOOKS_LOG, logLine);
+      process.stderr.write(logLine);
+    } catch (error) {
+      // Silent fail on log errors
+    }
+  }
+}
+
+// =============================================================================
+// MAIN
+// =============================================================================
+
+let input = '';
+process.stdin.setEncoding('utf8');
+process.stdin.on('data', (chunk) => { input += chunk; });
+
+process.stdin.on('end', async () => {
+  let payload;
+  try {
+    payload = input ? JSON.parse(input) : {};
+  } catch (error) {
+    debugLog('Stop: JSON parse error', { error: error.message });
+    console.log(JSON.stringify({ continue: true, suppressOutput: true }));
+    process.exit(0);
+  }
+
+  const { cwd } = payload;
+  const project = cwd ? getProjectName(cwd) : 'unknown';
+
+  // Immediately clear activity flag for UI indicator
+  const activityFlagPath = path.join(process.env.HOME || '', '.claude-mem', 'activity.flag');
+  try {
+    fs.writeFileSync(activityFlagPath, JSON.stringify({ active: false, timestamp: Date.now() }));
+  } catch (error) {
+    // Silent fail - non-critical
+  }
+
+  // Return immediately with async mode
+  console.log(JSON.stringify({ async: true, asyncTimeout: 180000 }));
+
+  try {
+    // Load SDK session info from database
+    const db = initializeDatabase();
+
+    const sessions = getActiveStreamingSessionsForProject(db, project);
+    if (!sessions || sessions.length === 0) {
+      debugLog('Stop: No streaming session found', { project });
+      db.close();
+      process.exit(0);
+    }
+
+    const sessionData = sessions[0];
+    const sdkSessionId = sessionData.sdk_session_id;
+    const claudeSessionId = sessionData.claude_session_id;
+
+    debugLog('Stop: Ending SDK session', { sdkSessionId, claudeSessionId });
+
+    // Build end message - SDK will call `claude-mem store-overview` and `chroma_add_documents`
+    const message = renderEndMessage({
+      project,
+      sessionId: claudeSessionId
+    });
+
+    // Send end message and wait for SDK to complete
+    const response = query({
+      prompt: message,
+      options: {
+        model: HOOK_CONFIG.sdk.model,
+        resume: sdkSessionId,
+        allowedTools: HOOK_CONFIG.sdk.allowedTools,
+        maxTokens: HOOK_CONFIG.sdk.maxTokensEnd,
+        cwd  // Must match where transcript was created
+      }
+    });
+
+    // Consume the response stream (wait for SDK to finish storing via CLI)
+    for await (const msg of response) {
+      if (msg.type === 'assistant' && msg.message?.content) {
+        for (const block of msg.message.content) {
+          if (block.type === 'tool_use') {
+            debugLog('Stop: SDK tool call', { tool: block.name });
+          }
+        }
+      }
+    }
+
+    debugLog('Stop: SDK session ended', { sdkSessionId });
+
+    // Delete SDK memories transcript from Claude Code UI
+    const sanitizedCwd = cwd.replace(/\//g, '-');
+    const projectsDir = path.join(process.env.HOME, '.claude', 'projects', sanitizedCwd);
+    const memoriesTranscriptPath = path.join(projectsDir, `${sdkSessionId}.jsonl`);
+
+    if (fs.existsSync(memoriesTranscriptPath)) {
+      fs.unlinkSync(memoriesTranscriptPath);
+      debugLog('Stop: Cleaned up memories transcript', { memoriesTranscriptPath });
+    }
+
+    // Mark session as completed in database
+    markStreamingSessionCompleted(db, sessionData.id);
+    debugLog('Stop: Session ended and marked complete', { project, sessionId: sessionData.id });
+
+    // Close database connection
+    db.close();
+  } catch (error) {
+    debugLog('Stop: Error ending session', { error: error.message });
+  }
+
+  // Exit cleanly after async processing completes
+  process.exit(0);
+});
@@ -0,0 +1,157 @@
+#!/usr/bin/env node
+
+/**
+ * User Prompt Submit Hook - Streaming SDK Version
+ *
+ * Starts a streaming SDK session that will process tool responses in real-time.
+ * Saves the SDK session ID for post-tool-use and stop hooks to resume.
+ */
+
+import path from 'path';
+import fs from 'fs';
+import { fileURLToPath } from 'url';
+import { query } from '@anthropic-ai/claude-agent-sdk';
+import { renderSystemPrompt, HOOK_CONFIG } from './shared/hook-prompt-renderer.js';
+import { getProjectName } from './shared/path-resolver.js';
+import { initializeDatabase, createStreamingSession, updateStreamingSession } from './shared/hook-helpers.js';
+
+const __filename = fileURLToPath(import.meta.url);
+const __dirname = path.dirname(__filename);
+
+const HOOKS_LOG = path.join(process.env.HOME || '', '.claude-mem', 'logs', 'hooks.log');
+
+function debugLog(message, data = {}) {
+  if (process.env.CLAUDE_MEM_DEBUG === 'true') {
+    const timestamp = new Date().toISOString();
+    const logLine = `[${timestamp}] HOOK DEBUG: ${message} ${JSON.stringify(data)}\n`;
+    try {
+      fs.appendFileSync(HOOKS_LOG, logLine);
+      process.stderr.write(logLine);
+    } catch (error) {
+      // Silent fail on log errors
+    }
+  }
+}
+
+// Removed: buildStreamingSystemPrompt function
+// Now using centralized config from hook-prompt-renderer.js
+
+// =============================================================================
+// MAIN
+// =============================================================================
+
+let input = '';
+process.stdin.setEncoding('utf8');
+process.stdin.on('data', (chunk) => { input += chunk; });
+
+process.stdin.on('end', async () => {
+  let payload;
+  try {
+    payload = input ? JSON.parse(input) : {};
+  } catch (error) {
+    debugLog('UserPromptSubmit: JSON parse error', { error: error.message });
+    console.log(JSON.stringify({ continue: true, suppressOutput: true }));
+    process.exit(0);
+  }
+
+  const { prompt, cwd, session_id, timestamp } = payload;
+  const project = cwd ? getProjectName(cwd) : 'unknown';
+  const date = timestamp ? timestamp.split('T')[0] : new Date().toISOString().split('T')[0];
+
+  debugLog('UserPromptSubmit: Starting streaming session', { project, session_id });
+
+  // Immediately signal activity start for UI indicator
+  const activityFlagPath = path.join(process.env.HOME || '', '.claude-mem', 'activity.flag');
+  try {
+    fs.writeFileSync(activityFlagPath, JSON.stringify({ active: true, project, timestamp: Date.now() }));
+  } catch (error) {
+    // Silent fail - non-critical
+  }
+
+  // Generate title and subtitle non-blocking
+  if (prompt && session_id && project) {
+    import('child_process').then(({ spawn }) => {
+      const titleProcess = spawn('claude-mem', [
+        'generate-title',
+        '--save',
+        '--project', project,
+        '--session', session_id,
+        prompt
+      ], {
+        stdio: 'ignore',
+        detached: true
+      });
+      titleProcess.unref();
+    }).catch(error => {
+      debugLog('UserPromptSubmit: Error spawning title generator', { error: error.message });
+    });
+  }
+
+  try {
+    // Initialize database and create session record FIRST
+    const db = initializeDatabase();
+
+    // Create session record immediately - this gives us a tracking ID
+    const sessionRecord = createStreamingSession(db, {
+      claude_session_id: session_id,
+      project,
+      user_prompt: prompt,
+      started_at: timestamp
+    });
+
+    debugLog('UserPromptSubmit: Created session record', {
+      internalId: sessionRecord.id,
+      claudeSessionId: session_id
+    });
+
+    // Build system prompt using centralized config
+    const systemPrompt = renderSystemPrompt({
+      project,
+      sessionId: session_id,
+      date,
+      userPrompt: prompt || ''
+    });
+
+    // Start SDK session using centralized config
+    const response = query({
+      prompt: systemPrompt,
+      options: {
+        model: HOOK_CONFIG.sdk.model,
+        allowedTools: HOOK_CONFIG.sdk.allowedTools,
+        maxTokens: HOOK_CONFIG.sdk.maxTokensSystem,
+        cwd  // SDK will save transcript in this directory
+      }
+    });
+
+    // Wait for session ID from init message and consume entire stream
+    let sdkSessionId = null;
+    for await (const message of response) {
+      if (message.type === 'system' && message.subtype === 'init') {
+        sdkSessionId = message.session_id;
+        debugLog('UserPromptSubmit: Got SDK session ID', { sdkSessionId });
+      }
+      // Don't break - consume entire stream so transcript gets written
+    }
+
+    if (sdkSessionId) {
+      // Update session record with SDK session ID
+      updateStreamingSession(db, sessionRecord.id, {
+        sdk_session_id: sdkSessionId
+      });
+
+      debugLog('UserPromptSubmit: SDK session started', {
+        internalId: sessionRecord.id,
+        sdkSessionId
+      });
+    }
+
+    // Close database connection
+    db.close();
+  } catch (error) {
+    debugLog('UserPromptSubmit: Error starting SDK session', { error: error.message });
+  }
+
+  // Return success to Claude Code
+  console.log(JSON.stringify({ continue: true, suppressOutput: true }));
+  process.exit(0);
+});
@@ -1,89 +0,0 @@
-#!/usr/bin/env node
-
-/**
- * Pre-Compact Hook for Claude Memory System
- * 
- * Updated to use the centralized PromptOrchestrator and HookTemplates system.
- * This hook validates the pre-compact request and executes compression using
- * standardized response templates for consistent Claude Code integration.
- */
-
-import { loadCliCommand } from './shared/config-loader.js';
-import { getLogsDir } from './shared/path-resolver.js';
-import { 
-  createHookResponse, 
-  executeCliCommand, 
-  validateHookPayload, 
-  debugLog 
-} from './shared/hook-helpers.js';
-
-
-// Set up stdin immediately
-process.stdin.setEncoding('utf8');
-process.stdin.resume(); // Explicitly enter flowing mode to prevent data loss
-
-
-// Read input from stdin
-let input = '';
-process.stdin.on('data', chunk => {
-  input += chunk;
-});
-
-process.stdin.on('end', async () => {
-  
-  try {
-    // Load CLI command inside try-catch to handle config errors properly
-    const cliCommand = loadCliCommand();
-    
-    const payload = JSON.parse(input);
-    debugLog('Pre-compact hook started', { payload });
-
-    // Validate payload using centralized validation
-    const validation = validateHookPayload(payload, 'PreCompact');
-    if (!validation.valid) {
-      const response = createHookResponse('PreCompact', false, { reason: validation.error });
-      debugLog('Validation failed', { response });
-      // Exit silently - validation failure is expected flow control
-      process.exit(0);
-    }
-
-    // Check for environment-based blocking conditions
-    if (payload.trigger === 'auto' && process.env.DISABLE_AUTO_COMPRESSION === 'true') {
-      const response = createHookResponse('PreCompact', false, { 
-        reason: 'Auto-compression disabled by configuration' 
-      });
-      debugLog('Auto-compression disabled', { response });
-      // Exit silently - disabled compression is expected flow control
-      process.exit(0);
-    }
-
-    // Execute compression using standardized CLI execution helper
-    debugLog('Executing compression command', { 
-      command: cliCommand, 
-      args: ['compress', payload.transcript_path] 
-    });
-    
-    const result = await executeCliCommand(cliCommand, ['compress', payload.transcript_path]);
-    
-    if (!result.success) {
-      const response = createHookResponse('PreCompact', false, { 
-        reason: `Compression failed: ${result.stderr || 'Unknown error'}` 
-      });
-      debugLog('Compression command failed', { stderr: result.stderr, response });
-      console.log(`claude-mem error: compression failed, see logs at ${getLogsDir()}`);
-      process.exit(1);  // Exit with error code for actual compression failure
-    }
-
-    // Success - exit silently (suppressOutput is true)
-    debugLog('Compression completed successfully');
-    process.exit(0);
-
-  } catch (error) {
-    const response = createHookResponse('PreCompact', false, { 
-      reason: `Hook execution error: ${error.message}` 
-    });
-    debugLog('Pre-compact hook error', { error: error.message, response });
-    console.log(`claude-mem error: hook failed, see logs at ${getLogsDir()}`);
-    process.exit(1);
-  }
-});
@@ -1,61 +0,0 @@
-#!/usr/bin/env node
-
-/**
- * Session End Hook - Handles session end events including /clear
- */
-
-import { loadCliCommand } from './shared/config-loader.js';
-import { getSettingsPath, getArchivesDir } from './shared/path-resolver.js';
-import { execSync } from 'child_process';
-import { existsSync, readFileSync } from 'fs';
-
-const cliCommand = loadCliCommand();
-
-// Check if save-on-clear is enabled
-function isSaveOnClearEnabled() {
-  const settingsPath = getSettingsPath();
-  if (existsSync(settingsPath)) {
-    try {
-      const settings = JSON.parse(readFileSync(settingsPath, 'utf8'));
-      return settings.saveMemoriesOnClear === true;
-    } catch (error) {
-      return false;
-    }
-  }
-  return false;
-}
-
-// Set up stdin immediately before any async operations
-process.stdin.setEncoding('utf8');
-process.stdin.resume(); // Explicitly enter flowing mode to prevent data loss
-
-// Read input
-let input = '';
-process.stdin.on('data', chunk => {
-  input += chunk;
-});
-
-process.stdin.on('end', async () => {
-  const data = JSON.parse(input);
-  
-  // Check if this is a clear event and save-on-clear is enabled
-  if (data.reason === 'clear' && isSaveOnClearEnabled()) {
-    console.error('🧠 Saving memories before clearing context...');
-    
-    try {
-      // Use the CLI to compress current transcript
-      execSync(`${cliCommand} compress --output ${getArchivesDir()}`, {
-        stdio: 'inherit',
-        env: { ...process.env, CLAUDE_MEM_SILENT: 'true' }
-      });
-      
-      console.error('✅ Memories saved successfully');
-    } catch (error) {
-      console.error('[session-end] Failed to save memories:', error.message);
-      // Don't block the clear operation if memory saving fails
-    }
-  }
-  
-  // Always continue
-  console.log(JSON.stringify({ continue: true }));
-});
@@ -1,170 +0,0 @@
-#!/usr/bin/env node
-
-/**
- * Session Start Hook - Load context when Claude Code starts
- * 
- * Updated to use the centralized PromptOrchestrator and HookTemplates system.
- * This hook loads previous session context using standardized formatting and
- * provides rich context messages for Claude Code integration.
- */
-
-import path from 'path';
-import { loadCliCommand } from './shared/config-loader.js';
-import { 
-  createHookResponse, 
-  formatSessionStartContext, 
-  executeCliCommand, 
-  parseContextData, 
-  validateHookPayload, 
-  debugLog 
-} from './shared/hook-helpers.js';
-
-const cliCommand = loadCliCommand();
-
-// Set up stdin immediately before any async operations
-process.stdin.setEncoding('utf8');
-process.stdin.resume(); // Explicitly enter flowing mode to prevent data loss
-
-// Read input from stdin
-let input = '';
-process.stdin.on('data', chunk => {
-  input += chunk;
-});
-
-process.stdin.on('end', async () => {
-  try {
-    const payload = JSON.parse(input);
-    debugLog('Session start hook started', { payload });
-
-    // Validate payload using centralized validation
-    const validation = validateHookPayload(payload, 'SessionStart');
-    if (!validation.valid) {
-      debugLog('Payload validation failed', { error: validation.error });
-      // For session start, continue even with invalid payload but log the error
-      const response = createHookResponse('SessionStart', false, { 
-        error: `Payload validation failed: ${validation.error}` 
-      });
-      console.log(JSON.stringify(response));
-      process.exit(0);
-    }
-
-    // Skip load-context when source is "resume" to avoid duplicate context
-    if (payload.source === 'resume') {
-      debugLog('Skipping load-context for resume source');
-      // Output valid JSON response with suppressOutput for resume
-      const response = createHookResponse('SessionStart', true);
-      console.log(JSON.stringify(response));
-      process.exit(0);
-    }
-
-    // Extract project name from current working directory
-    const projectName = path.basename(process.cwd());
-
-    // Load context using standardized CLI execution helper
-    const contextResult = await executeCliCommand(cliCommand, [
-      'load-context', 
-      '--format', 'session-start', 
-      '--project', projectName
-    ]);
-
-    if (!contextResult.success) {
-      debugLog('Context loading failed', { stderr: contextResult.stderr });
-      // Don't fail the session start, just provide error context
-      const response = createHookResponse('SessionStart', false, {
-        error: contextResult.stderr || 'Failed to load context'
-      });
-      console.log(JSON.stringify(response));
-      process.exit(0);
-    }
-
-    const rawContext = contextResult.stdout;
-    debugLog('Raw context loaded', { contextLength: rawContext.length });
-
-    // Check if the output is actually an error message (starts with ❌)
-    if (rawContext && rawContext.trim().startsWith('❌')) {
-      debugLog('Detected error message in stdout', { rawContext });
-      // Extract the clean error message without the emoji and format
-      const errorMatch = rawContext.match(/❌\s*[^:]+:\s*([^\n]+)(?:\n\n💡\s*(.+))?/);
-      let errorMsg = 'No previous memories found';
-      let suggestion = '';
-      
-      if (errorMatch) {
-        errorMsg = errorMatch[1] || errorMsg;
-        suggestion = errorMatch[2] || '';
-      }
-      
-      // Create a clean response without duplicating the error formatting
-      const response = createHookResponse('SessionStart', false, {
-        error: errorMsg + (suggestion ? `. ${suggestion}` : '')
-      });
-      console.log(JSON.stringify(response));
-      process.exit(0);
-    }
-
-    if (!rawContext || !rawContext.trim()) {
-      debugLog('No context available, creating empty response');
-      // No context available - use standardized empty response
-      const response = createHookResponse('SessionStart', true);
-      console.log(JSON.stringify(response));
-      process.exit(0);
-    }
-
-    // Parse context data and format using centralized templates
-    const contextData = parseContextData(rawContext);
-    contextData.projectName = projectName;
-    
-    // If we have raw context (not structured data), use it directly
-    let formattedContext;
-    if (contextData.rawContext) {
-      formattedContext = contextData.rawContext;
-    } else {
-      // Use standardized formatting for structured context
-      formattedContext = formatSessionStartContext(contextData);
-    }
-
-    debugLog('Context formatted successfully', { 
-      memoryCount: contextData.memoryCount,
-      hasStructuredData: !contextData.rawContext 
-    });
-
-    // Create standardized session start response using HookTemplates
-    const response = createHookResponse('SessionStart', true, { 
-      context: formattedContext 
-    });
-    
-    console.log(JSON.stringify(response));
-    process.exit(0);
-
-  } catch (error) {
-    debugLog('Session start hook error', { error: error.message });
-    // Even on error, continue the session with error information
-    const response = createHookResponse('SessionStart', false, { 
-      error: `Hook execution error: ${error.message}` 
-    });
-    console.log(JSON.stringify(response));
-    process.exit(0);
-  }
-});
-
-/**
- * Extracts project name from transcript path
- * @param {string} transcriptPath - Path to transcript file
- * @returns {string|null} Extracted project name or null
- */
-function extractProjectName(transcriptPath) {
-  if (!transcriptPath) return null;
-  
-  // Look for project pattern: /path/to/PROJECT_NAME/.claude/
-  // Need to get PROJECT_NAME, not the parent directory
-  const parts = transcriptPath.split(path.sep);
-  const claudeIndex = parts.indexOf('.claude');
-  
-  if (claudeIndex > 0) {
-    // Get the directory immediately before .claude
-    return parts[claudeIndex - 1];
-  }
-  
-  // Fall back to directory containing the transcript
-  const dir = path.dirname(transcriptPath);
-  return path.basename(dir);
-}
@@ -1,63 +0,0 @@
-#!/usr/bin/env node
-
-/**
- * Path resolver utility for Claude Memory hooks
- * Provides proper path handling using environment variables
- */
-
-import { join } from 'path';
-import { homedir } from 'os';
-
-/**
- * Gets the base data directory for claude-mem
- * @returns {string} Data directory path
- */
-export function getDataDir() {
-  return process.env.CLAUDE_MEM_DATA_DIR || join(homedir(), '.claude-mem');
-}
-
-/**
- * Gets the settings file path
- * @returns {string} Settings file path
- */
-export function getSettingsPath() {
-  return join(getDataDir(), 'settings.json');
-}
-
-/**
- * Gets the archives directory path
- * @returns {string} Archives directory path
- */
-export function getArchivesDir() {
-  return process.env.CLAUDE_MEM_ARCHIVES_DIR || join(getDataDir(), 'archives');
-}
-
-/**
- * Gets the logs directory path
- * @returns {string} Logs directory path
- */
-export function getLogsDir() {
-  return process.env.CLAUDE_MEM_LOGS_DIR || join(getDataDir(), 'logs');
-}
-
-/**
- * Gets the compact flag file path
- * @returns {string} Compact flag file path
- */
-export function getCompactFlagPath() {
-  return join(getDataDir(), '.compact-running');
-}
-
-/**
- * Gets all common paths used by hooks
- * @returns {Object} Object containing all common paths
- */
-export function getPaths() {
-  return {
-    dataDir: getDataDir(),
-    settingsPath: getSettingsPath(),
-    archivesDir: getArchivesDir(),
-    logsDir: getLogsDir(),
-    compactFlagPath: getCompactFlagPath()
-  };
-}
@@ -1,10 +1,10 @@
 {
  "name": "claude-mem",
-  "version": "3.7.1",
+  "version": "3.9.16",
  "description": "Memory compression system for Claude Code - persist context across sessions",
  "keywords": [
    "claude",
-    "claude-code",
+    "claude-agent-sdk",
    "mcp",
    "memory",
    "compression",
@@ -36,24 +36,22 @@
    "claude-mem": "./dist/claude-mem.min.js"
  },
  "dependencies": {
-    "@anthropic-ai/claude-code": "^1.0.88",
+    "@anthropic-ai/claude-agent-sdk": "^0.1.0",
    "@clack/prompts": "^0.11.0",
-    "@modelcontextprotocol/sdk": "^0.5.0",
    "boxen": "^8.0.1",
    "chalk": "^5.6.0",
-    "chromadb": "^3.0.14",
    "commander": "^14.0.0",
    "glob": "^11.0.3",
    "gradient-string": "^3.0.0",
-    "handlebars": "^4.7.8",
-    "oh-my-logo": "^0.3.2"
+    "handlebars": "^4.7.8"
  },
  "files": [
    "dist",
-    "hooks",
+    "hook-templates",
    "commands",
    "src",
    ".mcp.json",
-    "CHANGELOG.md"
+    "CHANGELOG.md",
+    "README_WINDOWS.md"
  ]
 }
@@ -7,20 +7,26 @@ import { Command } from 'commander';
 import { PACKAGE_NAME, PACKAGE_VERSION, PACKAGE_DESCRIPTION } from '../shared/config.js';

 // Import command handlers
-import { compress } from '../commands/compress.js';
 import { install } from '../commands/install.js';
 import { uninstall } from '../commands/uninstall.js';
 import { status } from '../commands/status.js';
 import { logs } from '../commands/logs.js';
 import { loadContext } from '../commands/load-context.js';
 import { trash } from '../commands/trash.js';
+import { viewTrash } from '../commands/trash-view.js';
+import { emptyTrash } from '../commands/trash-empty.js';
 import { restore } from '../commands/restore.js';
-import { save } from '../commands/save.js';
 import { changelog } from '../commands/changelog.js';
-// Cloud functionality disabled - incomplete setup
-// import { cloudCommand } from '../commands/cloud.js';
-import { importHistory } from '../commands/import-history.js';
-import { TranscriptCompressor } from '../core/compression/TranscriptCompressor.js';
+import { doctor } from '../commands/doctor.js';
+import { storeMemory } from '../commands/store-memory.js';
+import { storeOverview } from '../commands/store-overview.js';
+import { updateSessionMetadata } from '../commands/update-session-metadata.js';
+import { generateTitle } from '../commands/generate-title.js';
+import {
+  executeChromaMCPTool,
+  loadChromaMCPTools,
+  generateCommandOptions
+} from '../commands/chroma-mcp.js';

 const program = new Command();
 // </Block> =======================================
@@ -35,19 +41,6 @@ program
 // </Block> =======================================

 // <Block> 1.3 ====================================
-// Compress Command Definition
-// Natural pattern: Define command with its options and handler
-// Compress command
-program
-  .command('compress [transcript]')
-  .description('Compress a Claude Code transcript into memory')
-  .option('--output <path>', 'Output directory for compressed files')
-  .option('--dry-run', 'Show what would be compressed without doing it')
-  .option('-v, --verbose', 'Show detailed output')
-  .action(compress);
-// </Block> =======================================
-
-// <Block> 1.4 ====================================
 // Install Command Definition
 // Natural pattern: Define command with its options and handler
 // Install command
@@ -86,6 +79,20 @@ program
  .command('status')
  .description('Check installation status of Claude Memory System')
  .action(status);
+
+// Doctor command
+program
+  .command('doctor')
+  .description('Run environment and pipeline diagnostics for rolling memory')
+  .option('--json', 'Output JSON instead of text')
+  .action(async (options: any) => {
+    try {
+      await doctor(options);
+    } catch (error: any) {
+      console.error(`doctor failed: ${error.message || error}`);
+      process.exitCode = 1;
+    }
+  });
 // </Block> =======================================

 // <Block> 1.7 ====================================
@@ -148,20 +155,14 @@ const trashCmd = program
 trashCmd
  .command('view')
  .description('View contents of trash bin')
-  .action(async () => {
-    const { viewTrash } = await import('../commands/trash-view.js');
-    await viewTrash();
-  });
+  .action(viewTrash);

 // Trash empty subcommand
 trashCmd
  .command('empty')
  .description('Permanently delete all files in trash')
  .option('-f, --force', 'Skip confirmation prompt')
-  .action(async (options: any) => {
-    const { emptyTrash } = await import('../commands/trash-empty.js');
-    await emptyTrash(options);
-  });
+  .action(emptyTrash);

 // Restore command
 program
@@ -170,15 +171,39 @@ program
  .action(restore);
 // </Block> =======================================

-// Cloud command
-// Cloud functionality disabled - incomplete setup
-// program.addCommand(cloudCommand);
-
-// Save command
+// Store memory command (for SDK streaming)
 program
-  .command('save <message>')
-  .description('Save a message to the memory system')
-  .action(save);
+  .command('store-memory')
+  .description('Store a memory to all storage layers (used by SDK)')
+  .requiredOption('--id <id>', 'Memory ID')
+  .requiredOption('--project <project>', 'Project name')
+  .requiredOption('--session <session>', 'Session ID')
+  .requiredOption('--date <date>', 'Date (YYYY-MM-DD)')
+  .requiredOption('--title <title>', 'Memory title (3-8 words)')
+  .requiredOption('--subtitle <subtitle>', 'Memory subtitle (max 24 words)')
+  .requiredOption('--facts <json>', 'Atomic facts as JSON array')
+  .option('--concepts <json>', 'Concept tags as JSON array')
+  .option('--files <json>', 'Files touched as JSON array')
+  .action(storeMemory);
+
+// Store overview command (for SDK streaming)
+program
+  .command('store-overview')
+  .description('Store a session overview (used by SDK)')
+  .requiredOption('--project <project>', 'Project name')
+  .requiredOption('--session <session>', 'Session ID')
+  .requiredOption('--content <content>', 'Overview content')
+  .action(storeOverview);
+
+// Update session metadata command (for SDK streaming)
+program
+  .command('update-session-metadata')
+  .description('Update session title and subtitle (used by SDK)')
+  .requiredOption('--project <project>', 'Project name')
+  .requiredOption('--session <session>', 'Session ID')
+  .requiredOption('--title <title>', 'Session title (3-6 words)')
+  .option('--subtitle <subtitle>', 'Session subtitle (max 20 words)')
+  .action(updateSessionMetadata);

 // Changelog command
 program
@@ -193,67 +218,58 @@ program
  .option('-v, --verbose', 'Show detailed output')
  .action(changelog);

-// Import History command
+// Generate title command
 program
-  .command('import-history')
-  .description('Import historical Claude Code conversations into memory')
-  .option('-v, --verbose', 'Show detailed output')
-  .option('-m, --multi', 'Enable multi-select mode (default is single-select)')
-  .action(importHistory);
-
-// Migrate Index command
-program
-  .command('migrate-index')
-  .description('Migrate JSONL index to SQLite database')
-  .option('--force', 'Force migration even if SQLite database already has data')
-  .option('--keep-jsonl', 'Keep original JSONL file (archive it by default)')
-  .action(async (options) => {
-    const { migrateIndex } = await import('../commands/migrate-index.js');
-    await migrateIndex(options);
-  });
-
-// <Block> 1.11 ===================================  
-// Hook Commands
-// Internal commands called by hook scripts
-program
-  .command('hook:pre-compact', { hidden: true })
-  .description('Internal pre-compact hook handler')
-  .action(async () => {
-    const { preCompactHook } = await import('../commands/hooks.js');
-    await preCompactHook();
-  });
-
-program
-  .command('hook:session-start', { hidden: true })
-  .description('Internal session-start hook handler')
-  .action(async () => {
-    const { sessionStartHook } = await import('../commands/hooks.js');
-    await sessionStartHook();
-  });
-
-program
-  .command('hook:session-end', { hidden: true })
-  .description('Internal session-end hook handler')
-  .action(async () => {
-    const { sessionEndHook } = await import('../commands/hooks.js');
-    await sessionEndHook();
-  });
+  .command('generate-title <prompt>')
+  .description('Generate a session title and subtitle from a prompt')
+  .option('--json', 'Output as JSON')
+  .option('--oneline', 'Output as single line (title - subtitle)')
+  .option('--session-id <id>', 'Claude session ID to update')
+  .option('--save', 'Save the generated title to the database (requires --session-id)')
+  .action(generateTitle);

 // </Block> =======================================

-// Debug command to show filtered output
-program
-  .command('debug-filter')
-  .description('Show filtered transcript output (first 5 messages)')
-  .argument('<transcript-path>', 'Path to transcript file')
-  .action((transcriptPath) => {
-    const compressor = new TranscriptCompressor();
-    compressor.showFilteredOutput(transcriptPath);
-  });
+// <Block> 1.12 ===================================
+// Dynamic Chroma MCP Commands
+// Natural pattern: Register all Chroma MCP tools as CLI commands
+try {
+  const chromaTools = loadChromaMCPTools();
+
+  for (const tool of chromaTools) {
+    const cmd = program
+      .command(tool.name)
+      .description(tool.description || `Execute ${tool.name} MCP tool`);
+
+    // Add options from tool schema
+    const options = generateCommandOptions(tool.inputSchema);
+    for (const opt of options) {
+      if (opt.required) {
+        cmd.requiredOption(opt.flag, opt.description);
+      } else {
+        cmd.option(opt.flag, opt.description);
+      }
+    }
+
+    // Set action handler
+    cmd.action(async (options: OptionValues) => {
+      await executeChromaMCPTool(tool.name, options);
+    });
+  }
+} catch (error) {
+  console.warn('Warning: Could not load Chroma MCP tools:', error);
+}
+// </Block> =======================================

 // <Block> 1.11 ===================================
 // CLI Execution
 // Natural pattern: After defining all commands, parse and execute
 // Parse arguments and execute
 program.parse();
-// </Block> =======================================
+// </Block> =======================================
+
+// <Block> 1.12 ===================================
+// Module Exports for Programmatic Use
+// Export database and utility classes for hooks and external consumers
+export { DatabaseManager, StreamingSessionStore, migrations, initializeDatabase, getDatabase } from '../services/sqlite/index.js';
+// </Block> =======================================
@@ -1,5 +1,5 @@
 import { OptionValues } from 'commander';
-import { query } from '@anthropic-ai/claude-code';
+import { query } from '@anthropic-ai/claude-agent-sdk';
 import fs from 'fs';
 import path from 'path';
 import { getClaudePath } from '../shared/settings.js';
@@ -0,0 +1,196 @@
+import { OptionValues } from 'commander';
+import ChromaMCPClient from '../../chroma-mcp-tools/chroma-mcp-client.js';
+import fs from 'fs';
+import path from 'path';
+import { fileURLToPath } from 'url';
+
+const __filename = fileURLToPath(import.meta.url);
+const __dirname = path.dirname(__filename);
+
+/**
+ * Generic Chroma MCP tool executor
+ * Dynamically calls any Chroma MCP tool with provided arguments
+ */
+export async function executeChromaMCPTool(toolName: string, options: OptionValues): Promise<void> {
+  const client = new ChromaMCPClient();
+
+  try {
+    await client.connect();
+
+    // Convert commander options to tool arguments
+    const toolArgs = convertOptionsToArgs(toolName, options);
+
+    // Call the MCP tool
+    const result = await client.callTool(toolName, toolArgs);
+
+    // Parse and format the result nicely
+    const formatted = formatMCPResult(result);
+    console.log(formatted);
+
+    await client.disconnect();
+    process.exit(0);
+  } catch (error: any) {
+    console.error(JSON.stringify({
+      success: false,
+      error: error.message || 'Unknown error calling MCP tool',
+      tool: toolName
+    }, null, 2));
+
+    await client.disconnect();
+    process.exit(1);
+  }
+}
+
+/**
+ * Format MCP tool result for clean CLI output
+ */
+function formatMCPResult(result: any): string {
+  // If result has content array (MCP protocol format)
+  if (result?.content && Array.isArray(result.content)) {
+    const textContent = result.content
+      .filter((item: any) => item.type === 'text')
+      .map((item: any) => item.text)
+      .join('\n');
+
+    // Try to parse as JSON for prettier output
+    try {
+      const parsed = JSON.parse(textContent);
+      return JSON.stringify(parsed, null, 2);
+    } catch {
+      // Not JSON, return as-is
+      return textContent;
+    }
+  }
+
+  // If result is already an object, pretty print it
+  if (typeof result === 'object') {
+    return JSON.stringify(result, null, 2);
+  }
+
+  // Fallback to string
+  return String(result);
+}
+
+/**
+ * Convert CLI options to MCP tool arguments
+ * Handles type conversion and array parsing
+ */
+function convertOptionsToArgs(toolName: string, options: OptionValues): Record<string, any> {
+  const args: Record<string, any> = {};
+
+  for (const [key, value] of Object.entries(options)) {
+    // Skip commander internal properties
+    if (key.startsWith('_') || typeof value === 'function') {
+      continue;
+    }
+
+    // Try to parse JSON strings
+    if (typeof value === 'string') {
+      try {
+        args[key] = JSON.parse(value);
+      } catch {
+        args[key] = value;
+      }
+    } else {
+      args[key] = value;
+    }
+  }
+
+  return args;
+}
+
+/**
+ * Load Chroma MCP tool definitions from JSON
+ */
+export function loadChromaMCPTools(): Array<{
+  name: string;
+  description: string;
+  inputSchema: any;
+}> {
+  // Try multiple path resolutions for dev vs production
+  const possiblePaths = [
+    path.join(__dirname, '../../chroma-mcp-tools/CHROMA_MCP_TOOLS.json'),
+    path.join(process.cwd(), 'chroma-mcp-tools/CHROMA_MCP_TOOLS.json'),
+    path.join(__dirname, '../chroma-mcp-tools/CHROMA_MCP_TOOLS.json')
+  ];
+
+  for (const toolsPath of possiblePaths) {
+    if (fs.existsSync(toolsPath)) {
+      const toolsJson = fs.readFileSync(toolsPath, 'utf-8');
+      return JSON.parse(toolsJson);
+    }
+  }
+
+  throw new Error('Could not find CHROMA_MCP_TOOLS.json');
+}
+
+/**
+ * Generate CLI command options from MCP tool schema
+ */
+export function generateCommandOptions(schema: any): Array<{
+  flag: string;
+  description: string;
+  required: boolean;
+  type: string;
+}> {
+  const options: Array<{
+    flag: string;
+    description: string;
+    required: boolean;
+    type: string;
+  }> = [];
+
+  if (!schema.properties) return options;
+
+  const required = schema.required || [];
+
+  for (const [propName, propSchema] of Object.entries(schema.properties)) {
+    const prop = propSchema as any;
+    const isRequired = required.includes(propName);
+
+    // Determine type
+    let type = 'string';
+    if (prop.type === 'integer' || prop.type === 'number') {
+      type = 'number';
+    } else if (prop.type === 'array') {
+      type = 'array';
+    } else if (prop.type === 'object') {
+      type = 'json';
+    } else if (prop.anyOf) {
+      // Handle nullable types
+      const nonNullType = prop.anyOf.find((t: any) => t.type !== 'null');
+      if (nonNullType?.type === 'integer' || nonNullType?.type === 'number') {
+        type = 'number';
+      } else if (nonNullType?.type === 'array') {
+        type = 'array';
+      } else if (nonNullType?.type === 'object') {
+        type = 'json';
+      }
+    }
+
+    // Build flag
+    const flag = isRequired
+      ? `--${propName} <${type}>`
+      : `--${propName} [${type}]`;
+
+    // Build description
+    let description = prop.title || propName;
+    if (prop.default !== undefined) {
+      description += ` (default: ${JSON.stringify(prop.default)})`;
+    }
+    if (type === 'array') {
+      description += ' (JSON array)';
+    } else if (type === 'json') {
+      description += ' (JSON object)';
+    }
+
+    options.push({
+      flag,
+      description,
+      required: isRequired,
+      type
+    });
+  }
+
+  return options;
+}
@@ -1,43 +0,0 @@
-import { OptionValues } from 'commander';
-import { basename, dirname } from 'path';
-import { 
-  createLoadingMessage, 
-  createCompletionMessage, 
-  createOperationSummary,
-  createUserFriendlyError 
-} from '../prompts/templates/context/ContextTemplates.js';
-
-export async function compress(transcript?: string, options: OptionValues = {}): Promise<void> {
-  console.log(createLoadingMessage('compressing'));
-  
-  if (!transcript) {
-    console.log(createUserFriendlyError('Compression', 'No transcript file provided', 'Please provide a path to a transcript file'));
-    return;
-  }
-
-  try {
-    const startTime = Date.now();
-    
-    // Import and run compression
-    const { TranscriptCompressor } = await import('../core/compression/TranscriptCompressor.js');
-    const compressor = new TranscriptCompressor({
-      verbose: options.verbose || false
-    });
-    
-    const sessionId = options.sessionId || basename(transcript, '.jsonl');
-    const archivePath = await compressor.compress(transcript, sessionId);
-    
-    const duration = Date.now() - startTime;
-    
-    console.log(createCompletionMessage('Compression', undefined, `Session archived as ${basename(archivePath)}`));
-    console.log(createOperationSummary('compress', { count: 1, duration, details: `Session: ${sessionId}` }));
-  } catch (error) {
-    const errorMessage = error instanceof Error ? error.message : String(error);
-    console.log(createUserFriendlyError(
-      'Compression', 
-      errorMessage, 
-      'Check that the transcript file exists and you have write permissions'
-    ));
-    throw error; // Re-throw to maintain existing error handling behavior
-  }
-}
@@ -0,0 +1,100 @@
+import { OptionValues } from 'commander';
+import fs from 'fs';
+import path from 'path';
+import { PathDiscovery } from '../services/path-discovery.js';
+import { createStores } from '../services/sqlite/index.js';
+import { rollingLog } from '../shared/rolling-log.js';
+
+type CheckStatus = 'pass' | 'fail' | 'warn';
+
+interface CheckResult {
+  name: string;
+  status: CheckStatus;
+  details?: string;
+}
+
+function printCheck(result: CheckResult): void {
+  const icon =
+    result.status === 'pass' ? '✅' : result.status === 'warn' ? '⚠️ ' : '❌';
+  const message = result.details ? `${result.name}: ${result.details}` : result.name;
+  console.log(`${icon} ${message}`);
+}
+
+export async function doctor(options: OptionValues = {}): Promise<void> {
+  const discovery = PathDiscovery.getInstance();
+  const checks: CheckResult[] = [];
+
+  // Data directory
+  try {
+    const dataDir = discovery.getDataDirectory();
+    if (!fs.existsSync(dataDir)) {
+      fs.mkdirSync(dataDir, { recursive: true });
+      checks.push({ name: `Data directory created at ${dataDir}`, status: 'warn' });
+    } else {
+      const stats = fs.statSync(dataDir);
+      let writable = false;
+      try {
+        fs.accessSync(dataDir, fs.constants.W_OK);
+        writable = true;
+      } catch {}
+      checks.push({
+        name: `Data directory ${dataDir}`,
+        status: stats.isDirectory() && writable ? 'pass' : 'fail',
+        details: stats.isDirectory() && writable ? 'accessible' : 'not writable'
+      });
+    }
+  } catch (error: any) {
+    checks.push({
+      name: 'Data directory',
+      status: 'fail',
+      details: error?.message || String(error)
+    });
+  }
+
+  // SQLite connectivity
+  let stores; // reuse for queue check
+  try {
+    stores = await createStores();
+    const sessionCount = stores.sessions.count();
+    checks.push({
+      name: 'SQLite database',
+      status: 'pass',
+      details: `${sessionCount} session${sessionCount === 1 ? '' : 's'} present`
+    });
+  } catch (error: any) {
+    checks.push({
+      name: 'SQLite database',
+      status: 'fail',
+      details: error?.message || String(error)
+    });
+  }
+
+  // Chroma connectivity
+  try {
+    const chromaDir = discovery.getChromaDirectory();
+    const chromaExists = fs.existsSync(chromaDir);
+    checks.push({
+      name: 'Chroma vector store',
+      status: chromaExists ? 'pass' : 'warn',
+      details: chromaExists ? `data dir ${path.resolve(chromaDir)}` : 'Not yet initialized'
+    });
+  } catch (error: any) {
+    checks.push({
+      name: 'Chroma vector store',
+      status: 'warn',
+      details: error?.message || 'Unable to check Chroma directory'
+    });
+  }
+
+  if (options.json) {
+    console.log(JSON.stringify({ checks }, null, 2));
+  } else {
+    console.log('claude-mem doctor');
+    console.log('=================');
+    checks.forEach(printCheck);
+  }
+
+  rollingLog('info', 'doctor run completed', {
+    status: checks.map((c) => c.status)
+  });
+}
@@ -0,0 +1,181 @@
+import { OptionValues } from 'commander';
+import { query } from '@anthropic-ai/claude-agent-sdk';
+import { getClaudePath } from '../shared/settings.js';
+import { DatabaseManager } from '../services/sqlite/Database.js';
+import { StreamingSessionStore } from '../services/sqlite/StreamingSessionStore.js';
+import { migrations } from '../services/sqlite/migrations.js';
+
+/**
+ * Generate a session title and subtitle from a user prompt
+ * CLI command that uses Agent SDK (like changelog.ts)
+ *
+ * Can be called in two modes:
+ * 1. Standalone: generate-title "user prompt" --json
+ * 2. With session: generate-title "user prompt" --session-id <id> --save
+ */
+export async function generateTitle(prompt: string, options: OptionValues): Promise<void> {
+  if (!prompt || prompt.trim().length === 0) {
+    console.error(JSON.stringify({
+      success: false,
+      error: 'Prompt is required'
+    }));
+    process.exit(1);
+  }
+
+  // If --session-id provided, validate that session exists
+  let streamingStore: StreamingSessionStore | null = null;
+  let sessionRecord = null;
+
+  if (options.sessionId) {
+    try {
+      const dbManager = DatabaseManager.getInstance();
+      for (const migration of migrations) {
+        dbManager.registerMigration(migration);
+      }
+      const db = await dbManager.initialize();
+      streamingStore = new StreamingSessionStore(db);
+
+      sessionRecord = streamingStore.getByClaudeSessionId(options.sessionId);
+      if (!sessionRecord) {
+        console.error(JSON.stringify({
+          success: false,
+          error: `Session not found: ${options.sessionId}`
+        }));
+        process.exit(1);
+      }
+    } catch (error: any) {
+      console.error(JSON.stringify({
+        success: false,
+        error: `Database error: ${error.message}`
+      }));
+      process.exit(1);
+    }
+  }
+
+  const systemPrompt = `You are a title and subtitle generator for claude-mem session metadata.
+
+Your job is to analyze a user's request and generate:
+1. A concise title (3-8 words)
+2. A one-sentence subtitle (max 20 words)
+
+TITLE GUIDELINES:
+- 3-8 words maximum
+- Scannable and clear
+- Captures the core action or topic
+- Professional and informative
+- Examples:
+  * "Dark Mode Implementation"
+  * "Authentication Bug Fix"
+  * "API Rate Limiting Setup"
+  * "React Component Refactoring"
+
+SUBTITLE GUIDELINES:
+- One sentence, max 20 words
+- Descriptive and specific
+- Focus on the outcome or benefit
+- Use active voice when possible
+- Examples:
+  * "Adding theme toggle and dark color scheme support to the application"
+  * "Resolving login timeout issue affecting user session persistence"
+  * "Implementing request throttling to prevent API quota exhaustion"
+
+OUTPUT FORMAT:
+You must output EXACTLY two lines:
+Line 1: Title only (no prefix, no quotes)
+Line 2: Subtitle only (no prefix, no quotes)
+
+EXAMPLE:
+
+User request: "Help me add dark mode to my app"
+
+Output:
+Dark Mode Implementation
+Adding theme toggle and dark color scheme support to the application
+
+USER REQUEST:
+${prompt}
+
+Now generate the title and subtitle (two lines exactly):`;
+
+  try {
+    const response = await query({
+      prompt: systemPrompt,
+      options: {
+        allowedTools: [],
+        pathToClaudeCodeExecutable: getClaudePath()
+      }
+    });
+
+    // Extract text from response (same pattern as changelog.ts)
+    let fullResponse = '';
+    if (response && typeof response === 'object' && Symbol.asyncIterator in response) {
+      for await (const message of response) {
+        if (message?.type === 'assistant' && message?.message?.content) {
+          const content = message.message.content;
+          if (typeof content === 'string') {
+            fullResponse += content;
+          } else if (Array.isArray(content)) {
+            for (const block of content) {
+              if (block.type === 'text' && block.text) {
+                fullResponse += block.text;
+              }
+            }
+          }
+        }
+      }
+    }
+
+    // Parse the response - expecting exactly 2 lines
+    const lines = fullResponse.trim().split('\n').filter(line => line.trim().length > 0);
+
+    if (lines.length < 2) {
+      console.error(JSON.stringify({
+        success: false,
+        error: 'Could not generate title and subtitle',
+        response: fullResponse
+      }));
+      process.exit(1);
+    }
+
+    const title = lines[0].trim();
+    const subtitle = lines[1].trim();
+
+    // If --save and we have a session, update the database
+    if (options.save && streamingStore && sessionRecord) {
+      try {
+        streamingStore.update(sessionRecord.id, {
+          title,
+          subtitle
+        });
+      } catch (error: any) {
+        console.error(JSON.stringify({
+          success: false,
+          error: `Failed to save title: ${error.message}`
+        }));
+        process.exit(1);
+      }
+    }
+
+    // Output format depends on options
+    if (options.json) {
+      console.log(JSON.stringify({
+        success: true,
+        title,
+        subtitle,
+        sessionId: sessionRecord?.claude_session_id
+      }, null, 2));
+    } else if (options.oneline) {
+      console.log(`${title} - ${subtitle}`);
+    } else {
+      console.log(title);
+      console.log(subtitle);
+    }
+
+  } catch (error: any) {
+    console.error(JSON.stringify({
+      success: false,
+      error: error.message || 'Unknown error generating title'
+    }));
+    process.exit(1);
+  }
+}
@@ -1,146 +0,0 @@
-/**
- * Hook command handlers for binary distribution
- * These execute the actual hook logic embedded in the binary
- */
-
-import { basename, sep } from 'path';
-import { compress } from './compress.js';
-import { loadContext } from './load-context.js';
-
-/**
- * Pre-compact hook handler
- * Runs compression on the Claude Code transcript
- */
-export async function preCompactHook(): Promise<void> {
-  try {
-    // Read hook data from stdin (Claude Code sends JSON)
-    let inputData = '';
-    
-    // Set up stdin to read data
-    process.stdin.setEncoding('utf8');
-    
-    // Collect all input data
-    for await (const chunk of process.stdin) {
-      inputData += chunk;
-    }
-    
-    // Parse the JSON input
-    let transcriptPath: string | undefined;
-    
-    if (inputData) {
-      try {
-        const hookData = JSON.parse(inputData);
-        transcriptPath = hookData.transcript_path;
-      } catch (parseError) {
-        // If JSON parsing fails, treat the input as a direct path
-        transcriptPath = inputData.trim();
-      }
-    }
-    
-    // Fallback to environment variable or command line argument
-    if (!transcriptPath) {
-      transcriptPath = process.env.TRANSCRIPT_PATH || process.argv[2];
-    }
-    
-    if (!transcriptPath) {
-      console.log('🗜️ Compressing session transcript...');
-      console.log('❌ No transcript path provided to pre-compact hook');
-      console.log('Hook data received:', inputData || 'none');
-      console.log('Environment TRANSCRIPT_PATH:', process.env.TRANSCRIPT_PATH || 'not set');
-      console.log('Command line args:', process.argv.slice(2));
-      return;
-    }
-    
-    // Run compression with the transcript path
-    await compress(transcriptPath, { dryRun: false });
-  } catch (error: any) {
-    console.error('Pre-compact hook failed:', error.message);
-    process.exit(1);
-  }
-}
-
-/**
- * Session-start hook handler
- * Loads context for the new session
- */
-export async function sessionStartHook(): Promise<void> {
-  try {
-    // Read hook data from stdin (Claude Code sends JSON)
-    let inputData = '';
-    
-    // Set up stdin to read data
-    process.stdin.setEncoding('utf8');
-    
-    // Collect all input data
-    for await (const chunk of process.stdin) {
-      inputData += chunk;
-    }
-    
-    // Parse the JSON input to get the current working directory
-    let project: string | undefined;
-    
-    if (inputData) {
-      try {
-        const hookData = JSON.parse(inputData);
-        // Extract project name from cwd if provided
-        if (hookData.cwd) {
-          project = basename(hookData.cwd);
-        }
-      } catch (parseError) {
-        // If JSON parsing fails, continue without project filtering
-        console.error('Failed to parse session-start hook data:', parseError);
-      }
-    }
-    
-    // If no project from hook data, try to get from current working directory
-    if (!project) {
-      project = basename(process.cwd());
-    }
-    
-    // Load context with session-start format and project filtering
-    await loadContext({ format: 'session-start', count: '10', project });
-  } catch (error: any) {
-    console.error('Session-start hook failed:', error.message);
-    process.exit(1);
-  }
-}
-
-/**
- * Session-end hook handler
- * Compresses session transcript when ending with /clear
- */
-export async function sessionEndHook(): Promise<void> {
-  try {
-    // Read hook data from stdin (Claude Code sends JSON)
-    let inputData = '';
-    
-    // Set up stdin to read data
-    process.stdin.setEncoding('utf8');
-    
-    // Collect all input data
-    for await (const chunk of process.stdin) {
-      inputData += chunk;
-    }
-    
-    // Parse the JSON input to check the reason for session end
-    if (inputData) {
-      try {
-        const hookData = JSON.parse(inputData);
-        
-        // If reason is "clear", compress the session transcript before it's deleted
-        if (hookData.reason === 'clear' && hookData.transcript_path) {
-          console.log('🗜️ Compressing current session before /clear...');
-          await compress(hookData.transcript_path, { dryRun: false });
-        }
-      } catch (parseError) {
-        // If JSON parsing fails, log but don't fail the hook
-        console.error('Failed to parse hook data:', parseError);
-      }
-    }
-    
-    console.log('Session ended successfully');
-  } catch (error: any) {
-    console.error('Session-end hook failed:', error.message);
-    process.exit(1);
-  }
-}
@@ -1,536 +0,0 @@
-#!/usr/bin/env node
-import * as p from '@clack/prompts';
-import path from 'path';
-import fs from 'fs';
-import os from 'os';
-import chalk from 'chalk';
-import { TranscriptCompressor } from '../core/compression/TranscriptCompressor.js';
-import { TitleGenerator, TitleGenerationRequest } from '../core/titles/TitleGenerator.js';
-import { getStorageProvider, needsMigration } from '../shared/storage.js';
-
-interface ConversationMetadata {
-  sessionId: string;
-  timestamp: string;
-  messageCount: number;
-  branch?: string;
-  cwd: string;
-  fileSize: number;
-}
-
-interface ConversationItem extends ConversationMetadata {
-  filePath: string;
-  projectName: string;
-  parsedDate: Date;
-  relativeDate: string;
-}
-
-function formatFileSize(bytes: number): string {
-  if (bytes < 1024) return `${bytes}B`;
-  if (bytes < 1024 * 1024) return `${(bytes / 1024).toFixed(1)}KB`;
-  return `${(bytes / (1024 * 1024)).toFixed(1)}MB`;
-}
-
-function formatRelativeDate(date: Date): string {
-  const now = new Date();
-  const diffMs = now.getTime() - date.getTime();
-  const diffMins = Math.floor(diffMs / 60000);
-  const diffHours = Math.floor(diffMs / 3600000);
-  const diffDays = Math.floor(diffMs / 86400000);
-  
-  if (diffMins < 1) return 'just now';
-  if (diffMins < 60) return `${diffMins}m ago`;
-  if (diffHours < 24) return `${diffHours}h ago`;
-  if (diffDays < 7) return `${diffDays}d ago`;
-  if (diffDays < 30) return `${Math.floor(diffDays / 7)}w ago`;
-  if (diffDays < 365) return `${Math.floor(diffDays / 30)}mo ago`;
-  return `${Math.floor(diffDays / 365)}y ago`;
-}
-
-function parseTimestamp(timestamp: string, fallbackPath: string): Date {
-  try {
-    const parsed = new Date(timestamp);
-    if (!isNaN(parsed.getTime())) return parsed;
-  } catch {}
-  
-  // Fallback: try to extract from filename
-  const match = fallbackPath.match(/(\d{4})-(\d{2})-(\d{2})_(\d{2})-(\d{2})-(\d{2})/);
-  if (match) {
-    const [_, year, month, day, hour, minute, second] = match;
-    return new Date(
-      parseInt(year), 
-      parseInt(month) - 1, 
-      parseInt(day),
-      parseInt(hour), 
-      parseInt(minute), 
-      parseInt(second)
-    );
-  }
-  
-  // Last resort: file stats
-  const stats = fs.statSync(fallbackPath);
-  return stats.mtime;
-}
-
-function extractFirstUserMessage(filePath: string): string {
-  try {
-    const content = fs.readFileSync(filePath, 'utf-8');
-    const lines = content.trim().split('\n').filter(Boolean);
-    
-    for (const line of lines) {
-      try {
-        const message = JSON.parse(line);
-        if (message.type === 'user' && message.message?.content) {
-          const messageContent = message.message.content;
-          if (Array.isArray(messageContent)) {
-            const textContent = messageContent
-              .filter(item => item.type === 'text')
-              .map(item => item.text)
-              .join(' ');
-            if (textContent.trim()) return textContent.trim();
-          } else if (typeof messageContent === 'string') {
-            return messageContent.trim();
-          }
-        }
-      } catch {}
-    }
-    
-    return 'Conversation'; // Fallback
-  } catch {
-    return 'Conversation'; // Fallback
-  }
-}
-
-async function loadImportedSessions(): Promise<Set<string>> {
-  try {
-    // Check if migration is needed and warn the user
-    if (await needsMigration()) {
-      console.warn('⚠️  JSONL to SQLite migration recommended. Run: claude-mem migrate-index');
-    }
-
-    const storage = await getStorageProvider();
-    
-    // Use storage provider to get all session IDs efficiently
-    return await storage.getAllSessionIds();
-    
-  } catch (error) {
-    console.warn('Failed to load imported sessions, proceeding with empty set:', error);
-    return new Set<string>();
-  }
-}
-
-async function scanConversations(): Promise<{ conversations: ConversationItem[]; skippedCount: number }> {
-  const claudeDir = path.join(os.homedir(), '.claude', 'projects');
-  
-  if (!fs.existsSync(claudeDir)) {
-    return { conversations: [], skippedCount: 0 };
-  }
-  
-  const projects = fs.readdirSync(claudeDir)
-    .filter(dir => fs.statSync(path.join(claudeDir, dir)).isDirectory());
-  
-  const conversations: ConversationItem[] = [];
-  const importedSessionIds = await loadImportedSessions();
-  let skippedCount = 0;
-  
-  for (const project of projects) {
-    const projectDir = path.join(claudeDir, project);
-    const files = fs.readdirSync(projectDir)
-      .filter(file => file.endsWith('.jsonl'))
-      .map(file => path.join(projectDir, file));
-    
-    for (const filePath of files) {
-      try {
-        const content = fs.readFileSync(filePath, 'utf-8');
-        const lines = content.trim().split('\n').filter(Boolean);
-        
-        // Parse first line for metadata
-        const firstLine = JSON.parse(lines[0]);
-        const messageCount = lines.length;
-        const stats = fs.statSync(filePath);
-        const fileSize = stats.size;
-        
-        const metadata: ConversationMetadata = {
-          sessionId: firstLine.sessionId || path.basename(filePath, '.jsonl'),
-          timestamp: firstLine.timestamp || stats.mtime.toISOString(),
-          messageCount,
-          branch: firstLine.branch,
-          cwd: firstLine.cwd || projectDir,
-          fileSize
-        };
-        
-        // Skip if already imported
-        if (importedSessionIds.has(metadata.sessionId)) {
-          skippedCount++;
-          continue;
-        }
-        
-        const projectName = path.basename(path.dirname(filePath));
-        const parsedDate = parseTimestamp(metadata.timestamp, filePath);
-        const relativeDate = formatRelativeDate(parsedDate);
-        
-        conversations.push({
-          filePath,
-          ...metadata,
-          projectName,
-          parsedDate,
-          relativeDate
-        });
-      } catch {}
-    }
-  }
-  
-  return { conversations, skippedCount };
-}
-
-export async function importHistory(options: { verbose?: boolean; multi?: boolean } = {}) {
-  console.clear();
-  
-  p.intro(chalk.bgCyan.black(' CLAUDE-MEM IMPORT '));
-  
-  const s = p.spinner();
-  s.start('Scanning conversation history');
-  
-  const { conversations, skippedCount } = await scanConversations();
-  
-  if (conversations.length === 0) {
-    s.stop('No new conversations found');
-    const message = skippedCount > 0 
-      ? `All ${skippedCount} conversation${skippedCount === 1 ? ' is' : 's are'} already imported.`
-      : 'No conversations found.';
-    p.outro(chalk.yellow(message));
-    return;
-  }
-  
-  // Sort by date (newest first)
-  conversations.sort((a, b) => b.parsedDate.getTime() - a.parsedDate.getTime());
-  
-  const statusMessage = skippedCount > 0
-    ? `Found ${conversations.length} new conversation${conversations.length === 1 ? '' : 's'} (${skippedCount} already imported)`
-    : `Found ${conversations.length} new conversation${conversations.length === 1 ? '' : 's'}`;
-  s.stop(statusMessage);
-  
-  // Group conversations by project for better organization
-  const projectGroups = conversations.reduce((acc, conv) => {
-    if (!acc[conv.projectName]) acc[conv.projectName] = [];
-    acc[conv.projectName].push(conv);
-    return acc;
-  }, {} as Record<string, ConversationItem[]>);
-  
-  // Create selection options
-  const importMode = await p.select({
-    message: 'How would you like to import?',
-    options: [
-      { value: 'browse', label: 'Browse by Project', hint: 'Select project then conversations' },
-      { value: 'project', label: 'Import Entire Project', hint: 'Select project and import all conversations' },
-      { value: 'recent', label: 'Recent Conversations', hint: 'Import most recent across all projects' },
-      { value: 'search', label: 'Search', hint: 'Search for specific conversations' }
-    ]
-  });
-  
-  if (p.isCancel(importMode)) {
-    p.cancel('Import cancelled');
-    return;
-  }
-  
-  let selectedConversations: ConversationItem[] = [];
-  
-  if (importMode === 'browse') {
-    // Project selection
-    const projectOptions = Object.entries(projectGroups)
-      .sort((a, b) => b[1][0].parsedDate.getTime() - a[1][0].parsedDate.getTime())
-      .map(([project, convs]) => ({
-        value: project,
-        label: project,
-        hint: `${convs.length} conversation${convs.length === 1 ? '' : 's'}, latest: ${convs[0].relativeDate}`
-      }));
-    
-    const selectedProject = await p.select({
-      message: 'Select a project',
-      options: projectOptions
-    });
-    
-    if (p.isCancel(selectedProject)) {
-      p.cancel('Import cancelled');
-      return;
-    }
-    
-    const projectConvs = projectGroups[selectedProject as string];
-    
-    // Ask about title generation
-    const generateTitles = await p.confirm({
-      message: 'Would you like to generate titles for easier browsing?',
-      initialValue: false
-    });
-    
-    if (p.isCancel(generateTitles)) {
-      p.cancel('Import cancelled');
-      return;
-    }
-    
-    if (generateTitles) {
-      await processTitleGeneration(projectConvs, selectedProject as string);
-    }
-    
-    // Conversation selection within project
-    const titleGenerator = new TitleGenerator();
-    const convOptions = projectConvs.map(conv => {
-      const title = titleGenerator.getTitleForSession(conv.sessionId);
-      const displayTitle = title ? `"${title}" • ` : '';
-      return {
-        value: conv.sessionId,
-        label: `${displayTitle}${conv.relativeDate} • ${conv.messageCount} messages • ${formatFileSize(conv.fileSize)}`,
-        hint: conv.branch ? `branch: ${conv.branch}` : undefined
-      };
-    });
-    
-    if (options.multi) {
-      const selected = await p.multiselect({
-        message: `Select conversations from ${selectedProject} (Space to select, Enter to confirm)`,
-        options: convOptions,
-        required: false
-      });
-      
-      if (p.isCancel(selected)) {
-        p.cancel('Import cancelled');
-        return;
-      }
-      
-      const selectedIds = selected as string[];
-      selectedConversations = projectConvs.filter(c => selectedIds.includes(c.sessionId));
-    } else {
-      // Single select with continuous import
-      let continueImporting = true;
-      const importedInSession = new Set<string>();
-      
-      while (continueImporting && projectConvs.length > importedInSession.size) {
-        const availableConvs = projectConvs.filter(c => !importedInSession.has(c.sessionId));
-        
-        if (availableConvs.length === 0) break;
-        
-        const titleGenerator = new TitleGenerator();
-        const convOptions = availableConvs.map(conv => {
-          const title = titleGenerator.getTitleForSession(conv.sessionId);
-          const displayTitle = title ? `"${title}" • ` : '';
-          return {
-            value: conv.sessionId,
-            label: `${displayTitle}${conv.relativeDate} • ${conv.messageCount} messages • ${formatFileSize(conv.fileSize)}`,
-            hint: conv.branch ? `branch: ${conv.branch}` : undefined
-          };
-        });
-        
-        const selected = await p.select({
-          message: `Select a conversation (${importedInSession.size}/${projectConvs.length} imported)`,
-          options: [
-            ...convOptions,
-            { value: 'done', label: '✅ Done importing', hint: 'Exit import mode' }
-          ]
-        });
-        
-        if (p.isCancel(selected) || selected === 'done') {
-          continueImporting = false;
-          break;
-        }
-        
-        const conv = availableConvs.find(c => c.sessionId === selected);
-        if (conv) {
-          selectedConversations = [conv];
-          await processImport(selectedConversations, options.verbose);
-          importedInSession.add(conv.sessionId);
-        }
-      }
-      
-      if (importedInSession.size > 0) {
-        p.outro(chalk.green(`✅ Imported ${importedInSession.size} conversation${importedInSession.size === 1 ? '' : 's'}`));
-      } else {
-        p.outro(chalk.yellow('No conversations imported'));
-      }
-      return;
-    }
-    
-  } else if (importMode === 'project') {
-    // Project selection for importing entire project
-    const projectOptions = Object.entries(projectGroups)
-      .sort((a, b) => b[1][0].parsedDate.getTime() - a[1][0].parsedDate.getTime())
-      .map(([project, convs]) => ({
-        value: project,
-        label: project,
-        hint: `${convs.length} conversation${convs.length === 1 ? '' : 's'}, latest: ${convs[0].relativeDate}`
-      }));
-    
-    const selectedProject = await p.select({
-      message: 'Select a project to import all conversations',
-      options: projectOptions
-    });
-    
-    if (p.isCancel(selectedProject)) {
-      p.cancel('Import cancelled');
-      return;
-    }
-    
-    const projectConvs = projectGroups[selectedProject as string];
-    
-    // Ask about title generation
-    const generateTitles = await p.confirm({
-      message: 'Would you like to generate titles for easier browsing?',
-      initialValue: false
-    });
-    
-    if (p.isCancel(generateTitles)) {
-      p.cancel('Import cancelled');
-      return;
-    }
-    
-    if (generateTitles) {
-      await processTitleGeneration(projectConvs, selectedProject as string);
-    }
-    
-    const confirm = await p.confirm({
-      message: `Import all ${projectConvs.length} conversation${projectConvs.length === 1 ? '' : 's'} from ${selectedProject}?`
-    });
-    
-    if (p.isCancel(confirm) || !confirm) {
-      p.cancel('Import cancelled');
-      return;
-    }
-    
-    selectedConversations = projectConvs;
-    
-  } else if (importMode === 'recent') {
-    const limit = await p.text({
-      message: 'How many recent conversations?',
-      placeholder: '10',
-      initialValue: '10',
-      validate: (value) => {
-        const num = parseInt(value);
-        if (isNaN(num) || num < 1) return 'Please enter a valid number';
-        if (num > conversations.length) return `Only ${conversations.length} available`;
-      }
-    });
-    
-    if (p.isCancel(limit)) {
-      p.cancel('Import cancelled');
-      return;
-    }
-    
-    const count = parseInt(limit as string);
-    selectedConversations = conversations.slice(0, count);
-    
-  } else if (importMode === 'search') {
-    const searchTerm = await p.text({
-      message: 'Search conversations (project name or session ID)',
-      placeholder: 'Enter search term'
-    });
-    
-    if (p.isCancel(searchTerm)) {
-      p.cancel('Import cancelled');
-      return;
-    }
-    
-    const term = (searchTerm as string).toLowerCase();
-    const matches = conversations.filter(c => 
-      c.projectName.toLowerCase().includes(term) ||
-      c.sessionId.toLowerCase().includes(term) ||
-      (c.branch && c.branch.toLowerCase().includes(term))
-    );
-    
-    if (matches.length === 0) {
-      p.outro(chalk.yellow('No matching conversations found'));
-      return;
-    }
-    
-    const titleGenerator = new TitleGenerator();
-    const matchOptions = matches.map(conv => {
-      const title = titleGenerator.getTitleForSession(conv.sessionId);
-      const displayTitle = title ? `"${title}" • ` : '';
-      return {
-        value: conv.sessionId,
-        label: `${displayTitle}${conv.projectName} • ${conv.relativeDate} • ${conv.messageCount} msgs`,
-        hint: formatFileSize(conv.fileSize)
-      };
-    });
-    
-    const selected = await p.multiselect({
-      message: `Found ${matches.length} matches. Select to import:`,
-      options: matchOptions,
-      required: false
-    });
-    
-    if (p.isCancel(selected)) {
-      p.cancel('Import cancelled');
-      return;
-    }
-    
-    const selectedIds = selected as string[];
-    selectedConversations = matches.filter(c => selectedIds.includes(c.sessionId));
-  }
-  
-  // Process the import
-  if (selectedConversations.length > 0) {
-    await processImport(selectedConversations, options.verbose);
-    p.outro(chalk.green(`✅ Successfully imported ${selectedConversations.length} conversation${selectedConversations.length === 1 ? '' : 's'}`));
-  } else {
-    p.outro(chalk.yellow('No conversations selected for import'));
-  }
-}
-
-async function processTitleGeneration(conversations: ConversationItem[], projectName: string) {
-  const titleGenerator = new TitleGenerator();
-  const existingTitles = titleGenerator.getExistingTitles();
-  
-  // Filter conversations that don't have titles yet
-  const conversationsNeedingTitles = conversations.filter(conv => !existingTitles.has(conv.sessionId));
-  
-  if (conversationsNeedingTitles.length === 0) {
-    p.note('All conversations already have titles!', 'Title Generation');
-    return;
-  }
-  
-  const s = p.spinner();
-  s.start(`Generating titles for ${conversationsNeedingTitles.length} conversations...`);
-  
-  const requests: TitleGenerationRequest[] = conversationsNeedingTitles.map(conv => ({
-    sessionId: conv.sessionId,
-    projectName: projectName,
-    firstMessage: extractFirstUserMessage(conv.filePath)
-  }));
-  
-  try {
-    await titleGenerator.batchGenerateTitles(requests);
-    s.stop(`✅ Generated ${conversationsNeedingTitles.length} titles`);
-  } catch (error) {
-    s.stop(`❌ Failed to generate titles`);
-    console.error(chalk.red(`Error: ${error}`));
-  }
-}
-
-async function processImport(conversations: ConversationItem[], verbose?: boolean) {
-  const s = p.spinner();
-  
-  for (let i = 0; i < conversations.length; i++) {
-    const conv = conversations[i];
-    const progress = conversations.length > 1 ? `[${i + 1}/${conversations.length}] ` : '';
-    
-    s.start(`${progress}Importing ${conv.projectName} (${conv.relativeDate})`);
-    
-    try {
-      // Extract project name from the conversation's cwd field
-      const projectName = path.basename(conv.cwd);
-      
-      // Use TranscriptCompressor to process
-      const compressor = new TranscriptCompressor();
-      await compressor.compress(conv.filePath, conv.sessionId, projectName);
-      
-      s.stop(`${progress}Imported ${conv.projectName} (${conv.messageCount} messages)`);
-      
-      if (verbose) {
-        p.note(`Session: ${conv.sessionId}\nSize: ${formatFileSize(conv.fileSize)}\nBranch: ${conv.branch || 'main'}`, 'Details');
-      }
-      
-    } catch (error) {
-      s.stop(`${progress}Failed to import ${conv.projectName}`);
-      if (verbose) {
-        console.error(chalk.red(`Error: ${error}`));
-      }
-    }
-  }
-}
@@ -2,7 +2,7 @@ import { OptionValues } from 'commander';
 import fs from 'fs';
 import { join } from 'path';
 import { PathDiscovery } from '../services/path-discovery.js';
-import { 
+import {
  createCompletionMessage,
  createContextualError,
  createUserFriendlyError,
@@ -10,7 +10,10 @@ import {
  outputSessionStartContent
 } from '../prompts/templates/context/ContextTemplates.js';
 import { getStorageProvider, needsMigration } from '../shared/storage.js';
-import { MemoryRow, OverviewRow, SessionRow } from '../services/sqlite/types.js';
+import { MemoryRow, OverviewRow } from '../services/sqlite/types.js';
+import { createStores } from '../services/sqlite/index.js';
+import { getRollingSettings } from '../shared/rolling-settings.js';
+import { rollingLog } from '../shared/rolling-log.js';

 interface TrashStatus {
  folderCount: number;
@@ -19,6 +22,45 @@ interface TrashStatus {
  isEmpty: boolean;
 }

+function formatDateHeader(date = new Date()): string {
+  return date.toLocaleString('en-US', {
+    weekday: 'long',
+    month: 'long',
+    day: 'numeric',
+    year: 'numeric',
+    hour: '2-digit',
+    minute: '2-digit',
+    timeZoneName: 'short'
+  });
+}
+
+function wordWrap(text: string, maxWidth: number, prefix: string): string {
+  const words = text.split(' ');
+  const lines: string[] = [];
+  let currentLine = prefix;
+  const continuationPrefix = ' '.repeat(prefix.length);
+
+  for (const word of words) {
+    const needsSpace = currentLine !== prefix && currentLine !== continuationPrefix;
+    const testLine = currentLine + (needsSpace ? ' ' : '') + word;
+
+    if (testLine.length <= maxWidth) {
+      currentLine = testLine;
+    } else {
+      if (currentLine.trim()) {
+        lines.push(currentLine);
+      }
+      currentLine = continuationPrefix + word;
+    }
+  }
+
+  if (currentLine.trim()) {
+    lines.push(currentLine);
+  }
+
+  return lines.join('\n');
+}
+
 function buildProjectMatcher(projectName: string): (value?: string) => boolean {
  const aliases = new Set<string>();
  aliases.add(projectName);
@@ -67,6 +109,124 @@ function getTrashStatus(): TrashStatus {
  return { folderCount, fileCount, totalSize, isEmpty: false };
 }

+async function renderRollingSessionStart(projectOverride?: string): Promise<void> {
+  const settings = getRollingSettings();
+
+  if (!settings.sessionStartEnabled) {
+    console.log('Rolling session-start output disabled in settings.');
+    rollingLog('info', 'session-start output skipped (disabled)', {
+      project: projectOverride
+    });
+    return;
+  }
+
+  const stores = await createStores();
+  const projectName = projectOverride || PathDiscovery.getCurrentProjectName();
+
+  // Get all overviews for this project (oldest to newest)
+  const allOverviews = stores.overviews.getAllForProject(projectName);
+
+  // Limit to last 10 overviews
+  const recentOverviews = allOverviews.slice(-10);
+
+  // If no data at all, show friendly message
+  if (recentOverviews.length === 0) {
+    console.log('===============================================================================');
+    console.log(`What's new | ${formatDateHeader()}`);
+    console.log('===============================================================================');
+    console.log('No previous sessions found for this project.');
+    console.log('Start working and claude-mem will automatically capture context for future sessions.');
+    console.log('===============================================================================');
+    const trashStatus = getTrashStatus();
+    if (!trashStatus.isEmpty) {
+      const formattedSize = formatSize(trashStatus.totalSize);
+      console.log(
+        `🗑️ Trash – ${trashStatus.folderCount} folders | ${trashStatus.fileCount} files | ${formattedSize} – use \`claude-mem restore\``
+      );
+      console.log('===============================================================================');
+    }
+    return;
+  }
+
+  // Output header
+  console.log('===============================================================================');
+  console.log(`What's new | ${formatDateHeader()}`);
+  console.log('===============================================================================');
+
+  // Output each overview with timestamp, memory names, and files touched (oldest to newest)
+  recentOverviews.forEach((overview) => {
+    const date = new Date(overview.created_at);
+    const year = date.getFullYear();
+    const month = String(date.getMonth() + 1).padStart(2, '0');
+    const day = String(date.getDate()).padStart(2, '0');
+    const hours = date.getHours();
+    const minutes = String(date.getMinutes()).padStart(2, '0');
+    const ampm = hours >= 12 ? 'PM' : 'AM';
+    const displayHours = hours % 12 || 12;
+
+    console.log(`[${year}-${month}-${day} at ${displayHours}:${minutes} ${ampm}]`);
+
+    // Get memories for this session to show titles, subtitles, files, and keywords
+    const sessionMemories = stores.memories.getBySessionId(overview.session_id);
+
+    // Extract memory titles and subtitles
+    const memories = sessionMemories
+      .map(m => ({ title: m.title, subtitle: m.subtitle }))
+      .filter(m => m.title);
+
+    // Extract unique files touched across all memories
+    const allFilesTouched = new Set<string>();
+    const allKeywords = new Set<string>();
+
+    sessionMemories.forEach(m => {
+      if (m.files_touched) {
+        try {
+          const files = JSON.parse(m.files_touched);
+          if (Array.isArray(files)) {
+            files.forEach(f => allFilesTouched.add(f));
+          }
+        } catch (e) {
+          // Skip invalid JSON
+        }
+      }
+
+      if (m.keywords) {
+        // Keywords are comma-separated
+        m.keywords.split(',').forEach(k => allKeywords.add(k.trim()));
+      }
+    });
+
+    console.log('');
+
+    // Always show overview content
+    console.log(wordWrap(overview.content, 80, ''));
+
+    // Display files touched if any
+    if (allFilesTouched.size > 0) {
+      console.log('');
+      console.log(wordWrap(`- ${Array.from(allFilesTouched).join(', ')}`, 80, ''));
+    }
+
+    // Display keywords/tags if any
+    if (allKeywords.size > 0) {
+      console.log('');
+      console.log(wordWrap(`Tags: ${Array.from(allKeywords).join(', ')}`, 80, ''));
+    }
+
+    console.log('');
+  });
+
+  console.log('===============================================================================');
+  const trashStatus = getTrashStatus();
+  if (!trashStatus.isEmpty) {
+    const formattedSize = formatSize(trashStatus.totalSize);
+    console.log(
+      `🗑️ Trash – ${trashStatus.folderCount} folders | ${trashStatus.fileCount} files | ${formattedSize} – use \`claude-mem restore\``
+    );
+    console.log('===============================================================================');
+  }
+}
+
 export async function loadContext(options: OptionValues = {}): Promise<void> {
  try {
    // Check if migration is needed and warn the user
@@ -84,7 +244,6 @@ export async function loadContext(options: OptionValues = {}): Promise<void> {
    // SQLite implementation - fetch data using storage provider
    let recentMemories: MemoryRow[] = [];
    let recentOverviews: OverviewRow[] = [];
-    let recentSessions: SessionRow[] = [];

    // Auto-detect current project for session-start format if no project specified
    let projectToUse = options.project;
@@ -92,14 +251,19 @@ export async function loadContext(options: OptionValues = {}): Promise<void> {
      projectToUse = PathDiscovery.getCurrentProjectName();
    }

+    if (options.format === 'session-start') {
+      await renderRollingSessionStart(projectToUse);
+      return;
+    }
+
+    const overviewLimit = options.format === 'json' ? 5 : 3;
+
    if (projectToUse) {
      recentMemories = await storage.getRecentMemoriesForProject(projectToUse, 10);
-      recentOverviews = await storage.getRecentOverviewsForProject(projectToUse, options.format === 'session-start' ? 5 : 3);
-      recentSessions = await storage.getRecentSessionsForProject(projectToUse, 5);
+      recentOverviews = await storage.getRecentOverviewsForProject(projectToUse, overviewLimit);
    } else {
      recentMemories = await storage.getRecentMemories(10);
-      recentOverviews = await storage.getRecentOverviews(options.format === 'session-start' ? 5 : 3);
-      recentSessions = await storage.getRecentSessions(5);
+      recentOverviews = await storage.getRecentOverviews(overviewLimit);
    }

    // Convert SQLite rows to JSONL format for compatibility with existing output functions
@@ -122,48 +286,12 @@ export async function loadContext(options: OptionValues = {}): Promise<void> {
      timestamp: row.created_at
    }));

-    const sessionsAsJSON = recentSessions.map(row => ({
-      type: 'session',
-      session_id: row.session_id,
-      project: row.project,
-      timestamp: row.created_at
-    }));
-
    // If no data found, show appropriate messages
-    if (memoriesAsJSON.length === 0 && overviewsAsJSON.length === 0 && sessionsAsJSON.length === 0) {
-      if (options.format === 'session-start') {
-        console.log(createContextualError('NO_MEMORIES', projectToUse || 'this project'));
-      }
+    if (memoriesAsJSON.length === 0 && overviewsAsJSON.length === 0) {
      return;
    }

-    // Use the same output logic as the original implementation
-    if (options.format === 'session-start') {
-      // Combine them for the display
-      const recentObjects = [...sessionsAsJSON, ...memoriesAsJSON, ...overviewsAsJSON];
-      
-      // Find most recent timestamp for last session info
-      let lastSessionTime = 'recently';
-      const timestamps = recentObjects
-        .map(obj => {
-          return obj.timestamp ? new Date(obj.timestamp) : null;
-        })
-        .filter(date => date !== null)
-        .sort((a, b) => b.getTime() - a.getTime());
-      
-      if (timestamps.length > 0) {
-        lastSessionTime = formatTimeAgo(timestamps[0]);
-      }
-
-      // Use dual-stream output for session start formatting
-      outputSessionStartContent({
-        projectName: projectToUse || 'your project',
-        memoryCount: memoriesAsJSON.length,
-        lastSessionTime,
-        recentObjects
-      });
-      
-    } else if (options.format === 'json') {
+    if (options.format === 'json') {
      // For JSON format, combine last 10 of each type
      const recentObjects = [...memoriesAsJSON, ...overviewsAsJSON];
      console.log(JSON.stringify(recentObjects));
@@ -189,7 +317,7 @@ export async function loadContext(options: OptionValues = {}): Promise<void> {
      const trashStatus = getTrashStatus();
      if (!trashStatus.isEmpty) {
        const formattedSize = formatSize(trashStatus.totalSize);
-        console.log(`🗑️  Trash – ${trashStatus.folderCount} folders | ${trashStatus.fileCount} files | ${formattedSize} – use \`$ claude-mem restore\``);
+        console.log(`🗑️  Trash – ${trashStatus.folderCount} folders | ${trashStatus.fileCount} files | ${formattedSize} – use \`claude-mem restore\``);
        console.log('');
      }
    }
@@ -276,10 +404,10 @@ async function loadContextFromJSONL(options: OptionValues = {}): Promise<void> {
  }

  if (options.format === 'session-start') {
-    // Get last 10 memories and last 5 overviews for session-start
+    // Get last 10 memories and last 10 overviews for session-start
    const recentMemories = filteredMemories.slice(-10);
-    const recentOverviews = filteredOverviews.slice(-5);
-    const recentSessions = filteredSessions.slice(-5);
+    const recentOverviews = filteredOverviews.slice(-10);
+    const recentSessions = filteredSessions.slice(-10);
    
    // Combine them for the display
    const recentObjects = [...recentSessions, ...recentMemories, ...recentOverviews];
@@ -1,300 +0,0 @@
-import { OptionValues } from 'commander';
-import fs from 'fs';
-import path from 'path';
-import { PathDiscovery } from '../services/path-discovery.js';
-import { 
-  createStores, 
-  SessionInput, 
-  MemoryInput, 
-  OverviewInput, 
-  DiagnosticInput, 
-  normalizeTimestamp 
-} from '../services/sqlite/index.js';
-
-interface MigrationStats {
-  totalLines: number;
-  skippedLines: number;
-  invalidJson: number;
-  sessionsCreated: number;
-  memoriesCreated: number;
-  overviewsCreated: number;
-  diagnosticsCreated: number;
-  orphanedOverviews: number;
-  orphanedMemories: number;
-}
-
-/**
- * Migrate claude-mem index from JSONL to SQLite
- */
-export async function migrateIndex(options: OptionValues = {}): Promise<void> {
-  const pathDiscovery = PathDiscovery.getInstance();
-  const indexPath = pathDiscovery.getIndexPath();
-  const backupPath = `${indexPath}.backup-${Date.now()}`;
-
-  console.log('🔄 Starting JSONL to SQLite migration...');
-  console.log(`📁 Index file: ${indexPath}`);
-
-  // Check if JSONL file exists
-  if (!fs.existsSync(indexPath)) {
-    console.log('ℹ️  No JSONL index file found - nothing to migrate');
-    return;
-  }
-
-  try {
-    // Initialize SQLite database and stores
-    console.log('🏗️  Initializing SQLite database...');
-    const stores = await createStores();
-    
-    // Check if we already have data in SQLite
-    const existingSessions = stores.sessions.count();
-    if (existingSessions > 0 && !options.force) {
-      console.log(`⚠️  SQLite database already contains ${existingSessions} sessions.`);
-      console.log('   Use --force to migrate anyway (will skip duplicates)');
-      return;
-    }
-
-    // Create backup of JSONL file
-    if (!options.keepJsonl) {
-      console.log(`💾 Creating backup: ${path.basename(backupPath)}`);
-      fs.copyFileSync(indexPath, backupPath);
-    }
-
-    // Read and parse JSONL file
-    console.log('📖 Reading JSONL index file...');
-    const content = fs.readFileSync(indexPath, 'utf-8');
-    const lines = content.trim().split('\n').filter(line => line.trim());
-
-    const stats: MigrationStats = {
-      totalLines: lines.length,
-      skippedLines: 0,
-      invalidJson: 0,
-      sessionsCreated: 0,
-      memoriesCreated: 0,
-      overviewsCreated: 0,
-      diagnosticsCreated: 0,
-      orphanedOverviews: 0,
-      orphanedMemories: 0
-    };
-
-    console.log(`📝 Processing ${stats.totalLines} lines...`);
-
-    // Parse all lines first
-    const records: any[] = [];
-    for (let i = 0; i < lines.length; i++) {
-      const line = lines[i];
-      
-      try {
-        // Skip lines that don't look like JSON
-        if (!line.trim().startsWith('{')) {
-          stats.skippedLines++;
-          continue;
-        }
-        
-        const record = JSON.parse(line);
-        if (record && typeof record === 'object') {
-          records.push({ ...record, _lineNumber: i + 1 });
-        } else {
-          stats.skippedLines++;
-        }
-      } catch (error) {
-        stats.invalidJson++;
-        console.warn(`⚠️  Invalid JSON at line ${i + 1}: ${line.substring(0, 50)}...`);
-      }
-    }
-
-    console.log(`✅ Parsed ${records.length} valid records`);
-
-    // Group records by type
-    const sessions = records.filter(r => r.type === 'session');
-    const memories = records.filter(r => r.type === 'memory');
-    const overviews = records.filter(r => r.type === 'overview');
-    const diagnostics = records.filter(r => r.type === 'diagnostic');
-    const unknown = records.filter(r => !['session', 'memory', 'overview', 'diagnostic'].includes(r.type));
-
-    if (unknown.length > 0) {
-      console.log(`⚠️  Found ${unknown.length} records with unknown types - will skip`);
-      stats.skippedLines += unknown.length;
-    }
-
-    // Create session tracking
-    const sessionIds = new Set(sessions.map(s => s.session_id));
-    const orphanedSessionIds = new Set();
-
-    // Migrate sessions first
-    console.log('💾 Migrating sessions...');
-    for (const sessionData of sessions) {
-      try {
-        const { isoString } = normalizeTimestamp(sessionData.timestamp);
-        
-        const sessionInput: SessionInput = {
-          session_id: sessionData.session_id,
-          project: sessionData.project || 'unknown',
-          created_at: isoString,
-          source: 'legacy-jsonl'
-        };
-
-        // Skip if session already exists (when using --force)
-        if (!stores.sessions.has(sessionInput.session_id)) {
-          stores.sessions.create(sessionInput);
-          stats.sessionsCreated++;
-        }
-      } catch (error) {
-        console.warn(`⚠️  Failed to migrate session ${sessionData.session_id}: ${error}`);
-      }
-    }
-
-    // Migrate memories
-    console.log('🧠 Migrating memories...');
-    for (const memoryData of memories) {
-      try {
-        const { isoString } = normalizeTimestamp(memoryData.timestamp);
-        
-        // Check if session exists, create orphaned session if needed
-        if (!sessionIds.has(memoryData.session_id)) {
-          if (!orphanedSessionIds.has(memoryData.session_id)) {
-            orphanedSessionIds.add(memoryData.session_id);
-            
-            const orphanedSession: SessionInput = {
-              session_id: memoryData.session_id,
-              project: memoryData.project || 'unknown',
-              created_at: isoString,
-              source: 'legacy-jsonl'
-            };
-
-            if (!stores.sessions.has(orphanedSession.session_id)) {
-              stores.sessions.create(orphanedSession);
-              stats.sessionsCreated++;
-              stats.orphanedMemories++;
-            }
-          }
-        }
-
-        const memoryInput: MemoryInput = {
-          session_id: memoryData.session_id,
-          text: memoryData.text || '',
-          document_id: memoryData.document_id,
-          keywords: memoryData.keywords,
-          created_at: isoString,
-          project: memoryData.project || 'unknown',
-          archive_basename: memoryData.archive,
-          origin: 'transcript'
-        };
-
-        // Skip duplicate document_ids
-        if (!memoryInput.document_id || !stores.memories.hasDocumentId(memoryInput.document_id)) {
-          stores.memories.create(memoryInput);
-          stats.memoriesCreated++;
-        }
-      } catch (error) {
-        console.warn(`⚠️  Failed to migrate memory ${memoryData.document_id}: ${error}`);
-      }
-    }
-
-    // Migrate overviews
-    console.log('📋 Migrating overviews...');
-    for (const overviewData of overviews) {
-      try {
-        const { isoString } = normalizeTimestamp(overviewData.timestamp);
-        
-        // Check if session exists, create orphaned session if needed
-        if (!sessionIds.has(overviewData.session_id)) {
-          if (!orphanedSessionIds.has(overviewData.session_id)) {
-            orphanedSessionIds.add(overviewData.session_id);
-            
-            const orphanedSession: SessionInput = {
-              session_id: overviewData.session_id,
-              project: overviewData.project || 'unknown',
-              created_at: isoString,
-              source: 'legacy-jsonl'
-            };
-
-            if (!stores.sessions.has(orphanedSession.session_id)) {
-              stores.sessions.create(orphanedSession);
-              stats.sessionsCreated++;
-              stats.orphanedOverviews++;
-            }
-          }
-        }
-
-        const overviewInput: OverviewInput = {
-          session_id: overviewData.session_id,
-          content: overviewData.content || '',
-          created_at: isoString,
-          project: overviewData.project || 'unknown',
-          origin: 'claude'
-        };
-
-        stores.overviews.upsert(overviewInput);
-        stats.overviewsCreated++;
-      } catch (error) {
-        console.warn(`⚠️  Failed to migrate overview ${overviewData.session_id}: ${error}`);
-      }
-    }
-
-    // Migrate diagnostics
-    console.log('🩺 Migrating diagnostics...');
-    for (const diagnosticData of diagnostics) {
-      try {
-        const { isoString } = normalizeTimestamp(diagnosticData.timestamp);
-        
-        const diagnosticInput: DiagnosticInput = {
-          session_id: diagnosticData.session_id,
-          message: diagnosticData.message || '',
-          severity: 'warn',
-          created_at: isoString,
-          project: diagnosticData.project || 'unknown',
-          origin: 'compressor'
-        };
-
-        stores.diagnostics.create(diagnosticInput);
-        stats.diagnosticsCreated++;
-      } catch (error) {
-        console.warn(`⚠️  Failed to migrate diagnostic: ${error}`);
-      }
-    }
-
-    // Print migration summary
-    console.log('\n✅ Migration completed successfully!');
-    console.log('\n📊 Migration Summary:');
-    console.log(`   Total lines processed: ${stats.totalLines}`);
-    console.log(`   Skipped lines: ${stats.skippedLines}`);
-    console.log(`   Invalid JSON lines: ${stats.invalidJson}`);
-    console.log(`   Sessions created: ${stats.sessionsCreated}`);
-    console.log(`   Memories created: ${stats.memoriesCreated}`);
-    console.log(`   Overviews created: ${stats.overviewsCreated}`);
-    console.log(`   Diagnostics created: ${stats.diagnosticsCreated}`);
-    
-    if (stats.orphanedOverviews > 0 || stats.orphanedMemories > 0) {
-      console.log(`   Orphaned records (sessions synthesized): ${stats.orphanedOverviews + stats.orphanedMemories}`);
-    }
-
-    // Archive or keep JSONL file
-    if (options.keepJsonl) {
-      console.log(`\n💾 Original JSONL file preserved: ${indexPath}`);
-      console.log(`   SQLite database is now the primary index`);
-    } else {
-      const archiveDir = path.join(pathDiscovery.getDataDirectory(), 'archive', 'legacy');
-      fs.mkdirSync(archiveDir, { recursive: true });
-      
-      const archivedPath = path.join(archiveDir, `claude-mem-index-${Date.now()}.jsonl`);
-      fs.renameSync(indexPath, archivedPath);
-      
-      console.log(`\n📦 Original JSONL file archived: ${path.basename(archivedPath)}`);
-      console.log(`   Backup available at: ${path.basename(backupPath)}`);
-    }
-
-    console.log('\n🎉 Migration complete! You can now use claude-mem with SQLite backend.');
-    console.log('   Run `claude-mem load-context` to verify the migration worked.');
-
-  } catch (error) {
-    console.error('\n❌ Migration failed:', error);
-    
-    // Restore backup if we created one
-    if (fs.existsSync(backupPath) && !fs.existsSync(indexPath)) {
-      console.log('🔄 Restoring backup...');
-      fs.renameSync(backupPath, indexPath);
-    }
-    
-    process.exit(1);
-  }
-}
@@ -1,90 +0,0 @@
-import { OptionValues } from 'commander';
-import { appendFileSync } from 'fs';
-import { PathDiscovery } from '../services/path-discovery.js';
-import { getStorageProvider, needsMigration } from '../shared/storage.js';
-
-/**
- * Generates a descriptive session ID from the message content
- * Takes first few meaningful words and creates a readable identifier
- */
-function generateSessionId(message: string): string {
-  // Remove punctuation and split into words
-  const words = message
-    .toLowerCase()
-    .replace(/[^\w\s]/g, ' ')
-    .split(/\s+/)
-    .filter(word => word.length > 2); // Skip short words like 'a', 'is', 'to'
-  
-  // Take first 3-4 meaningful words, max 30 chars
-  const sessionWords = words.slice(0, 4).join('-');
-  const truncated = sessionWords.length > 30 ? sessionWords.substring(0, 27) + '...' : sessionWords;
-  
-  // Add timestamp suffix to ensure uniqueness
-  const timestamp = new Date().toISOString().substring(11, 19).replace(/:/g, '');
-  
-  return `${truncated}-${timestamp}`;
-}
-
-/**
- * Save command - stores a message using the configured storage provider
- */
-export async function save(message: string, options: OptionValues = {}): Promise<void> {
-  // Debug: Log what we receive
-  appendFileSync('/Users/alexnewman/.claude-mem/save-debug.log',
-    `[${new Date().toISOString()}] Received message: "${message}" (type: ${typeof message}, length: ${message?.length})\n`,
-    'utf8');
-
-  if (!message || message.trim() === '') {
-    console.error('Error: Message is required');
-    process.exit(1);
-  }
-
-  const timestamp = new Date().toISOString();
-  const projectName = PathDiscovery.getCurrentProjectName();
-  const sessionId = generateSessionId(message);
-  const documentId = `${projectName}_${sessionId}_overview`;
-
-  try {
-    // Check if migration is needed
-    if (await needsMigration()) {
-      console.warn('⚠️  JSONL to SQLite migration recommended. Run: claude-mem migrate-index');
-    }
-
-    // Get storage provider (SQLite preferred, JSONL fallback)
-    const storage = await getStorageProvider();
-
-    // Ensure session exists or create it
-    if (!await storage.hasSession(sessionId)) {
-      await storage.createSession({
-        session_id: sessionId,
-        project: projectName,
-        created_at: timestamp,
-        source: 'save'
-      });
-    }
-
-    // Upsert the overview
-    await storage.upsertOverview({
-      session_id: sessionId,
-      content: message,
-      created_at: timestamp,
-      project: projectName,
-      origin: 'manual'
-    });
-
-    // Return JSON response for hook compatibility
-    console.log(JSON.stringify({
-      success: true,
-      document_id: documentId,
-      session_id: sessionId,
-      project: projectName,
-      timestamp: timestamp,
-      backend: storage.backend,
-      suppressOutput: true
-    }));
-    
-  } catch (error) {
-    console.error('Error saving message:', error);
-    process.exit(1);
-  }
-}
@@ -3,6 +3,9 @@ import { join, resolve, dirname } from 'path';
 import { execSync } from 'child_process';
 import { fileURLToPath } from 'url';
 import { PathDiscovery } from '../services/path-discovery.js';
+import { DatabaseManager } from '../services/sqlite/Database.js';
+import { SessionStore } from '../services/sqlite/SessionStore.js';
+import chalk from 'chalk';

 const __dirname = dirname(fileURLToPath(import.meta.url));

@@ -10,13 +13,14 @@ export async function status(): Promise<void> {
    console.log('🔍 Claude Memory System Status Check');
    console.log('=====================================\n');
    
-    console.log('📂 Installed Hook Scripts:');
+    console.log('📂 Runtime Hook Scripts (installed from hook-templates/):');
    const pathDiscovery = PathDiscovery.getInstance();
-    const claudeMemHooksDir = pathDiscovery.getHooksDirectory();
-    const preCompactScript = join(claudeMemHooksDir, 'pre-compact.js');
-    const sessionStartScript = join(claudeMemHooksDir, 'session-start.js');
-    const sessionEndScript = join(claudeMemHooksDir, 'session-end.js');
-    
+    const runtimeHooksDir = pathDiscovery.getHooksDirectory();
+    const sessionStartScript = join(runtimeHooksDir, 'session-start.js');
+    const stopScript = join(runtimeHooksDir, 'stop.js');
+    const userPromptScript = join(runtimeHooksDir, 'user-prompt-submit.js');
+    const postToolScript = join(runtimeHooksDir, 'post-tool-use.js');
+
    const checkScript = (path: string, name: string) => {
      if (existsSync(path)) {
        console.log(`  ✅ ${name}: Found at ${path}`);
@@ -24,10 +28,11 @@ export async function status(): Promise<void> {
        console.log(`  ❌ ${name}: Not found at ${path}`);
      }
    };
-    
-    checkScript(preCompactScript, 'pre-compact.js');
+
    checkScript(sessionStartScript, 'session-start.js');
-    checkScript(sessionEndScript, 'session-end.js');
+    checkScript(stopScript, 'stop.js');
+    checkScript(userPromptScript, 'user-prompt-submit.js');
+    checkScript(postToolScript, 'post-tool-use.js');
    
    console.log('');
    
@@ -43,28 +48,35 @@ export async function status(): Promise<void> {
      
      try {
        const settings = JSON.parse(readFileSync(path, 'utf8'));
-        
-        const hasPreCompact = settings.hooks?.PreCompact?.some((matcher: any) => 
-          matcher.hooks?.some((hook: any) => 
-            hook.command?.includes('pre-compact.js') || hook.command?.includes('claude-mem')
-          )
-        );
-        
+
        const hasSessionStart = settings.hooks?.SessionStart?.some((matcher: any) =>
-          matcher.hooks?.some((hook: any) => 
+          matcher.hooks?.some((hook: any) =>
            hook.command?.includes('session-start.js') || hook.command?.includes('claude-mem')
          )
        );
-        
-        const hasSessionEnd = settings.hooks?.SessionEnd?.some((matcher: any) =>
-          matcher.hooks?.some((hook: any) => 
-            hook.command?.includes('session-end.js') || hook.command?.includes('claude-mem')
+
+        const hasStop = settings.hooks?.Stop?.some((matcher: any) =>
+          matcher.hooks?.some((hook: any) =>
+            hook.command?.includes('stop.js') || hook.command?.includes('claude-mem')
          )
        );
-        
-        console.log(`     PreCompact: ${hasPreCompact ? '✅' : '❌'}`);
+
+        const hasUserPrompt = settings.hooks?.UserPromptSubmit?.some((matcher: any) =>
+          matcher.hooks?.some((hook: any) =>
+            hook.command?.includes('user-prompt-submit.js') || hook.command?.includes('claude-mem')
+          )
+        );
+
+        const hasPostTool = settings.hooks?.PostToolUse?.some((matcher: any) =>
+          matcher.hooks?.some((hook: any) =>
+            hook.command?.includes('post-tool-use.js') || hook.command?.includes('claude-mem')
+          )
+        );
+
        console.log(`     SessionStart: ${hasSessionStart ? '✅' : '❌'}`);
-        console.log(`     SessionEnd: ${hasSessionEnd ? '✅' : '❌'}`);
+        console.log(`     Stop: ${hasStop ? '✅' : '❌'}`);
+        console.log(`     UserPromptSubmit: ${hasUserPrompt ? '✅' : '❌'}`);
+        console.log(`     PostToolUse: ${hasPostTool ? '✅' : '❌'}`);
        
      } catch (error: any) {
        console.log(`     ⚠️  Could not parse settings`);
@@ -136,7 +148,32 @@ export async function status(): Promise<void> {
    console.log('  ✅ Storage backend: Chroma MCP');
    console.log(`  📍 Data location: ${pathDiscovery.getChromaDirectory()}`);
    console.log('  🔍 Features: Vector search, semantic similarity, document storage');
-    
+
+    console.log('');
+
+    console.log('🤖 Claude Agent SDK Sessions:');
+    try {
+      const dbManager = DatabaseManager.getInstance();
+      await dbManager.initialize();
+      const sessionStore = new SessionStore();
+      const sessions = sessionStore.getAll();
+
+      if (sessions.length === 0) {
+        console.log(chalk.gray('  No active sessions'));
+      } else {
+        const activeCount = sessions.filter(s => {
+          const daysSinceUse = (Date.now() - s.last_used_epoch) / (1000 * 60 * 60 * 24);
+          return daysSinceUse < 7;
+        }).length;
+
+        console.log(`  📊 Total sessions: ${sessions.length}`);
+        console.log(`  ✅ Active (< 7 days): ${activeCount}`);
+        console.log(chalk.dim(`  💡 View details: claude-mem sessions list`));
+      }
+    } catch (error) {
+      console.log(chalk.gray('  ⚠️  Could not load session info'));
+    }
+
    console.log('');
    
    console.log('📊 Summary:');
@@ -149,15 +186,15 @@ export async function status(): Promise<void> {
    try {
      if (existsSync(globalPath)) {
        const settings = JSON.parse(readFileSync(globalPath, 'utf8'));
-        if (settings.hooks?.PreCompact || settings.hooks?.SessionStart || settings.hooks?.SessionEnd) {
+        if (settings.hooks?.SessionStart || settings.hooks?.Stop || settings.hooks?.PostToolUse) {
          isInstalled = true;
          installLocation = 'Global';
        }
      }
-      
+
      if (existsSync(projectPath)) {
        const settings = JSON.parse(readFileSync(projectPath, 'utf8'));
-        if (settings.hooks?.PreCompact || settings.hooks?.SessionStart || settings.hooks?.SessionEnd) {
+        if (settings.hooks?.SessionStart || settings.hooks?.Stop || settings.hooks?.PostToolUse) {
          isInstalled = true;
          installLocation = installLocation === 'Global' ? 'Global + Project' : 'Project';
        }
@@ -0,0 +1,154 @@
+import { OptionValues } from 'commander';
+import { spawnSync } from 'child_process';
+import { createStores } from '../services/sqlite/index.js';
+
+/**
+ * Store a memory to all three storage layers
+ * Called by SDK via bash during streaming memory capture
+ */
+export async function storeMemory(options: OptionValues): Promise<void> {
+  const { id, project, session, date, title, subtitle, facts, concepts, files } = options;
+
+  // Validate required fields
+  if (!id || !project || !session || !date) {
+    console.error('Error: All fields required: --id, --project, --session, --date');
+    process.exit(1);
+  }
+
+  // Validate hierarchical fields (required for v2 format)
+  if (!title || !subtitle || !facts) {
+    console.error('Error: Hierarchical format required: --title, --subtitle, --facts');
+    process.exit(1);
+  }
+
+  try {
+    const stores = await createStores();
+    const timestamp = new Date().toISOString();
+
+    // Ensure session exists
+    const sessionExists = await stores.sessions.has(session);
+    if (!sessionExists) {
+      await stores.sessions.create({
+        session_id: session,
+        project,
+        created_at: timestamp,
+        source: 'save'
+      });
+    }
+
+    // Parse JSON arrays if provided as strings
+    let factsArray: string | undefined;
+    let conceptsArray: string | undefined;
+    let filesArray: string | undefined;
+
+    try {
+      factsArray = facts ? JSON.stringify(JSON.parse(facts)) : undefined;
+    } catch (e) {
+      factsArray = facts; // Store as-is if not valid JSON
+    }
+
+    try {
+      conceptsArray = concepts ? JSON.stringify(JSON.parse(concepts)) : undefined;
+    } catch (e) {
+      conceptsArray = concepts; // Store as-is if not valid JSON
+    }
+
+    try {
+      filesArray = files ? JSON.stringify(JSON.parse(files)) : undefined;
+    } catch (e) {
+      filesArray = files; // Store as-is if not valid JSON
+    }
+
+    // Layer 1: SQLite Memory Index
+    const memoryExists = stores.memories.hasDocumentId(id);
+    if (!memoryExists) {
+      stores.memories.create({
+        document_id: id,
+        text: '', // Deprecated: hierarchical fields replace narrative text
+        keywords: '',
+        session_id: session,
+        project,
+        created_at: timestamp,
+        origin: 'streaming-sdk',
+        // Hierarchical fields (v2)
+        title: title || undefined,
+        subtitle: subtitle || undefined,
+        facts: factsArray,
+        concepts: conceptsArray,
+        files_touched: filesArray
+      });
+    }
+
+    // Layer 2: ChromaDB - Store hierarchical memory
+    if (factsArray) {
+      const factsJson = JSON.parse(factsArray);
+      const conceptsJson = conceptsArray ? JSON.parse(conceptsArray) : [];
+      const filesJson = filesArray ? JSON.parse(filesArray) : [];
+
+      // Store each atomic fact as a separate ChromaDB document
+      factsJson.forEach((fact: string, idx: number) => {
+        spawnSync('claude-mem', [
+          'chroma_add_documents',
+          '--collection_name', 'claude_memories',
+          '--documents', JSON.stringify([fact]),
+          '--ids', JSON.stringify([`${id}_fact_${String(idx).padStart(3, '0')}`]),
+          '--metadatas', JSON.stringify([{
+            type: 'fact',
+            parent_id: id,
+            fact_index: idx,
+            title,
+            subtitle,
+            project,
+            session_id: session,
+            created_at: timestamp,
+            created_at_epoch: Date.parse(timestamp),
+            keywords: '',
+            concepts: JSON.stringify(conceptsJson),
+            files_touched: JSON.stringify(filesJson),
+            origin: 'streaming-sdk'
+          }])
+        ]);
+      });
+
+      // Store full narrative with hierarchical metadata
+      spawnSync('claude-mem', [
+        'chroma_add_documents',
+        '--collection_name', 'claude_memories',
+        '--documents', JSON.stringify([`${title}\n${subtitle}\n\n${factsJson.join('\n')}`]),
+        '--ids', JSON.stringify([id]),
+        '--metadatas', JSON.stringify([{
+          type: 'narrative',
+          title,
+          subtitle,
+          facts_count: factsJson.length,
+          project,
+          session_id: session,
+          created_at: timestamp,
+          created_at_epoch: Date.parse(timestamp),
+          keywords: '',
+          concepts: JSON.stringify(conceptsJson),
+          files_touched: JSON.stringify(filesJson),
+          origin: 'streaming-sdk'
+        }])
+      ]);
+    }
+
+    // Success output (SDK will see this)
+    console.log(JSON.stringify({
+      success: true,
+      memory_id: id,
+      project,
+      session,
+      date,
+      timestamp,
+      hierarchical: !!(title && subtitle && facts)
+    }));
+
+  } catch (error: any) {
+    console.error(JSON.stringify({
+      success: false,
+      error: error.message || 'Unknown error storing memory'
+    }));
+    process.exit(1);
+  }
+}
@@ -0,0 +1,45 @@
+import { OptionValues } from 'commander';
+import { createStores } from '../services/sqlite/index.js';
+
+/**
+ * Store a session overview
+ * Called by SDK via bash at session end
+ */
+export async function storeOverview(options: OptionValues): Promise<void> {
+  const { project, session, content } = options;
+
+  // Validate required fields
+  if (!project || !session || !content) {
+    console.error('Error: All fields required: --project, --session, --content');
+    process.exit(1);
+  }
+
+  try {
+    const stores = await createStores();
+    const timestamp = new Date().toISOString();
+
+    // Create one overview per session (rolling log architecture)
+    stores.overviews.upsert({
+      session_id: session,
+      content,
+      created_at: timestamp,
+      project,
+      origin: 'streaming-sdk'
+    });
+
+    // Success output (SDK will see this)
+    console.log(JSON.stringify({
+      success: true,
+      project,
+      session,
+      timestamp
+    }));
+
+  } catch (error: any) {
+    console.error(JSON.stringify({
+      success: false,
+      error: error.message || 'Unknown error storing overview'
+    }));
+    process.exit(1);
+  }
+}
@@ -1,8 +1,69 @@
 import { OptionValues } from 'commander';
 import { readFileSync, writeFileSync, existsSync } from 'fs';
 import { join } from 'path';
+import { homedir } from 'os';
 import { PathDiscovery } from '../services/path-discovery.js';

+async function removeSmartTrashAlias(): Promise<boolean> {
+  const homeDir = homedir();
+  const shellConfigs = [
+    join(homeDir, '.bashrc'),
+    join(homeDir, '.zshrc'),
+    join(homeDir, '.bash_profile')
+  ];
+
+  const aliasLine = 'alias rm="claude-mem trash"';
+  // Handle both variations of the comment line
+  const commentPatterns = [
+    '# claude-mem smart trash alias',
+    '# claude-mem trash bin alias'
+  ];
+  let removedFromAny = false;
+
+  for (const configPath of shellConfigs) {
+    if (!existsSync(configPath)) continue;
+
+    let content = readFileSync(configPath, 'utf8');
+
+    // Check if alias exists
+    if (!content.includes(aliasLine)) {
+      continue; // Not configured in this file
+    }
+
+    // Remove the alias and its comment
+    const lines = content.split('\n');
+    const filteredLines = lines.filter((line, index) => {
+      // Skip the alias line
+      if (line.trim() === aliasLine) return false;
+      // Skip any claude-mem comment line if it's right before the alias
+      for (const commentPattern of commentPatterns) {
+        if (line.trim() === commentPattern &&
+            index + 1 < lines.length &&
+            lines[index + 1].trim() === aliasLine) {
+          return false;
+        }
+      }
+      return true;
+    });
+
+    const newContent = filteredLines.join('\n');
+
+    // Only write if content actually changed
+    if (newContent !== content) {
+      // Create backup
+      const backupPath = configPath + '.backup.' + Date.now();
+      writeFileSync(backupPath, content);
+
+      // Write updated content
+      writeFileSync(configPath, newContent);
+      console.log(`✅ Removed Smart Trash alias from ${configPath.replace(homeDir, '~')}`);
+      removedFromAny = true;
+    }
+  }
+
+  return removedFromAny;
+}
+
 export async function uninstall(options: OptionValues = {}): Promise<void> {
  console.log('🔄 Uninstalling Claude Memory System hooks...');
  
@@ -26,10 +87,10 @@ export async function uninstall(options: OptionValues = {}): Promise<void> {
  }
  
  const pathDiscovery = PathDiscovery.getInstance();
-  const claudeMemHooksDir = pathDiscovery.getHooksDirectory();
-  const preCompactScript = join(claudeMemHooksDir, 'pre-compact.js');
-  const sessionStartScript = join(claudeMemHooksDir, 'session-start.js');
-  const sessionEndScript = join(claudeMemHooksDir, 'session-end.js');
+  const runtimeHooksDir = pathDiscovery.getHooksDirectory();
+  const preCompactScript = join(runtimeHooksDir, 'pre-compact.js');
+  const sessionStartScript = join(runtimeHooksDir, 'session-start.js');
+  const sessionEndScript = join(runtimeHooksDir, 'session-end.js');
  
  let removedCount = 0;
  
@@ -39,95 +100,99 @@ export async function uninstall(options: OptionValues = {}): Promise<void> {
      continue;
    }
    
-    try {
-      const content = readFileSync(location.path, 'utf8');
-      const settings = JSON.parse(content);
-      
-      if (!settings.hooks) {
-        console.log(`⏭️  No hooks configured in ${location.name} settings`);
-        continue;
+    const content = readFileSync(location.path, 'utf8');
+    const settings = JSON.parse(content);
+
+    if (!settings.hooks) {
+      console.log(`⏭️  No hooks configured in ${location.name} settings`);
+      continue;
+    }
+
+    let modified = false;
+
+    if (settings.hooks.PreCompact) {
+      const filteredPreCompact = settings.hooks.PreCompact.filter((matcher: any) =>
+        !matcher.hooks?.some((hook: any) =>
+          hook.command === preCompactScript ||
+          hook.command?.includes('pre-compact.js') ||
+          hook.command?.includes('claude-mem')
+        )
+      );
+
+      if (filteredPreCompact.length !== settings.hooks.PreCompact.length) {
+        settings.hooks.PreCompact = filteredPreCompact.length ? filteredPreCompact : undefined;
+        modified = true;
+        console.log(`✅ Removed PreCompact hook from ${location.name} settings`);
      }
-      
-      let modified = false;
-      
-      if (settings.hooks.PreCompact) {
-        const filteredPreCompact = settings.hooks.PreCompact.filter((matcher: any) => 
-          !matcher.hooks?.some((hook: any) => 
-            hook.command === preCompactScript ||
-            hook.command?.includes('pre-compact.js') ||
-            hook.command?.includes('claude-mem')
-          )
-        );
-        
-        if (filteredPreCompact.length !== settings.hooks.PreCompact.length) {
-          settings.hooks.PreCompact = filteredPreCompact.length ? filteredPreCompact : undefined;
-          modified = true;
-          console.log(`✅ Removed PreCompact hook from ${location.name} settings`);
-        }
+    }
+
+    if (settings.hooks.SessionStart) {
+      const filteredSessionStart = settings.hooks.SessionStart.filter((matcher: any) =>
+        !matcher.hooks?.some((hook: any) =>
+          hook.command === sessionStartScript ||
+          hook.command?.includes('session-start.js') ||
+          hook.command?.includes('claude-mem')
+        )
+      );
+
+      if (filteredSessionStart.length !== settings.hooks.SessionStart.length) {
+        settings.hooks.SessionStart = filteredSessionStart.length ? filteredSessionStart : undefined;
+        modified = true;
+        console.log(`✅ Removed SessionStart hook from ${location.name} settings`);
      }
-      
-      if (settings.hooks.SessionStart) {
-        const filteredSessionStart = settings.hooks.SessionStart.filter((matcher: any) => 
-          !matcher.hooks?.some((hook: any) => 
-            hook.command === sessionStartScript ||
-            hook.command?.includes('session-start.js') ||
-            hook.command?.includes('claude-mem')
-          )
-        );
-        
-        if (filteredSessionStart.length !== settings.hooks.SessionStart.length) {
-          settings.hooks.SessionStart = filteredSessionStart.length ? filteredSessionStart : undefined;
-          modified = true;
-          console.log(`✅ Removed SessionStart hook from ${location.name} settings`);
-        }
+    }
+
+    if (settings.hooks.SessionEnd) {
+      const filteredSessionEnd = settings.hooks.SessionEnd.filter((matcher: any) =>
+        !matcher.hooks?.some((hook: any) =>
+          hook.command === sessionEndScript ||
+          hook.command?.includes('session-end.js') ||
+          hook.command?.includes('claude-mem')
+        )
+      );
+
+      if (filteredSessionEnd.length !== settings.hooks.SessionEnd.length) {
+        settings.hooks.SessionEnd = filteredSessionEnd.length ? filteredSessionEnd : undefined;
+        modified = true;
+        console.log(`✅ Removed SessionEnd hook from ${location.name} settings`);
      }
-      
-      if (settings.hooks.SessionEnd) {
-        const filteredSessionEnd = settings.hooks.SessionEnd.filter((matcher: any) => 
-          !matcher.hooks?.some((hook: any) => 
-            hook.command === sessionEndScript ||
-            hook.command?.includes('session-end.js') ||
-            hook.command?.includes('claude-mem')
-          )
-        );
-        
-        if (filteredSessionEnd.length !== settings.hooks.SessionEnd.length) {
-          settings.hooks.SessionEnd = filteredSessionEnd.length ? filteredSessionEnd : undefined;
-          modified = true;
-          console.log(`✅ Removed SessionEnd hook from ${location.name} settings`);
-        }
-      }
-      
-      if (settings.hooks.PreCompact === undefined) delete settings.hooks.PreCompact;
-      if (settings.hooks.SessionStart === undefined) delete settings.hooks.SessionStart;
-      if (settings.hooks.SessionEnd === undefined) delete settings.hooks.SessionEnd;
-      if (!Object.keys(settings.hooks).length) delete settings.hooks;
-      
-      if (modified) {
-        const backupPath = location.path + '.backup.' + Date.now();
-        writeFileSync(backupPath, content);
-        console.log(`📋 Created backup: ${backupPath}`);
-        
-        writeFileSync(location.path, JSON.stringify(settings, null, 2));
-        removedCount++;
-        console.log(`✅ Updated ${location.name} settings: ${location.path}`);
-      } else {
-        console.log(`ℹ️  No Claude Memory System hooks found in ${location.name} settings`);
-      }
-      
-    } catch (error: any) {
-      console.log(`⚠️  Could not process ${location.name} settings: ${error.message}`);
+    }
+
+    if (settings.hooks.PreCompact === undefined) delete settings.hooks.PreCompact;
+    if (settings.hooks.SessionStart === undefined) delete settings.hooks.SessionStart;
+    if (settings.hooks.SessionEnd === undefined) delete settings.hooks.SessionEnd;
+    if (!Object.keys(settings.hooks).length) delete settings.hooks;
+
+    if (modified) {
+      const backupPath = location.path + '.backup.' + Date.now();
+      writeFileSync(backupPath, content);
+      console.log(`📋 Created backup: ${backupPath}`);
+
+      writeFileSync(location.path, JSON.stringify(settings, null, 2));
+      removedCount++;
+      console.log(`✅ Updated ${location.name} settings: ${location.path}`);
+    } else {
+      console.log(`ℹ️  No Claude Memory System hooks found in ${location.name} settings`);
    }
  }
  
+  // Remove Smart Trash alias from shell configs
+  const removedAlias = await removeSmartTrashAlias();
+
  console.log('');
-  if (removedCount > 0) {
+  if (removedCount > 0 || removedAlias) {
    console.log('✨ Uninstallation complete!');
-    console.log('The Claude Memory System hooks have been removed from your settings.');
+    if (removedCount > 0) {
+      console.log('The Claude Memory System hooks have been removed from your settings.');
+    }
+    if (removedAlias) {
+      console.log('The Smart Trash alias has been removed from your shell configuration.');
+      console.log('⚠️  Restart your terminal for the alias removal to take effect.');
+    }
    console.log('');
    console.log('Note: Your compressed transcripts and archives are preserved.');
    console.log('To reinstall: claude-mem install');
  } else {
-    console.log('ℹ️  No Claude Memory System hooks were found to remove.');
+    console.log('ℹ️  No Claude Memory System hooks or aliases were found to remove.');
  }
 }
@@ -0,0 +1,80 @@
+import { OptionValues } from 'commander';
+import path from 'path';
+import fs from 'fs';
+
+const SESSION_DIR = path.join(process.env.HOME || '', '.claude-mem', 'sessions');
+
+/**
+ * Update session metadata (title/subtitle) in the streaming session JSON file
+ * Called by SDK when generating session title at the start
+ */
+export async function updateSessionMetadata(options: OptionValues): Promise<void> {
+  const { project, session, title, subtitle } = options;
+
+  // Validate required fields
+  if (!project || !session) {
+    console.error(JSON.stringify({
+      success: false,
+      error: 'Missing required fields: --project, --session'
+    }));
+    process.exit(1);
+  }
+
+  if (!title) {
+    console.error(JSON.stringify({
+      success: false,
+      error: 'Missing required field: --title'
+    }));
+    process.exit(1);
+  }
+
+  try {
+    // Load existing session file
+    const sessionFile = path.join(SESSION_DIR, `${project}_streaming.json`);
+
+    if (!fs.existsSync(sessionFile)) {
+      console.error(JSON.stringify({
+        success: false,
+        error: `Session file not found: ${sessionFile}`
+      }));
+      process.exit(1);
+    }
+
+    let sessionData: any = {};
+    try {
+      sessionData = JSON.parse(fs.readFileSync(sessionFile, 'utf8'));
+    } catch (e) {
+      console.error(JSON.stringify({
+        success: false,
+        error: 'Failed to parse session file'
+      }));
+      process.exit(1);
+    }
+
+    // Update metadata
+    sessionData.promptTitle = title;
+    if (subtitle) {
+      sessionData.promptSubtitle = subtitle;
+    }
+    sessionData.updatedAt = new Date().toISOString();
+
+    // Write back to file
+    fs.writeFileSync(sessionFile, JSON.stringify(sessionData, null, 2));
+
+    // Output success
+    console.log(JSON.stringify({
+      success: true,
+      title,
+      subtitle: subtitle || null,
+      project,
+      session
+    }));
+
+  } catch (error: any) {
+    console.error(JSON.stringify({
+      success: false,
+      error: error.message || 'Unknown error updating session metadata'
+    }));
+    process.exit(1);
+  }
+}
@@ -1,92 +1,9 @@
 /**
 * Claude Memory System - Core Constants
- * 
- * This file contains core application constants, CLI messages, 
- * configuration templates, and infrastructure-related constants.
+ *
+ * This file contains debug logging templates used throughout the application.
 */

-// =============================================================================
-// CONFIGURATION TEMPLATES
-// =============================================================================
-
-/**
- * Hook configuration templates for Claude settings
- */
-export const HOOK_CONFIG_TEMPLATES = {
-  PRE_COMPACT: (scriptPath: string) => ({
-    pattern: "*",
-    hooks: [{
-      type: "command",
-      command: scriptPath,
-      timeout: 180
-    }]
-  }),
-
-  SESSION_START: (scriptPath: string) => ({
-    pattern: "*", 
-    hooks: [{
-      type: "command",
-      command: scriptPath,
-      timeout: 30
-    }]
-  }),
-
-  SESSION_END: (scriptPath: string) => ({
-    pattern: "*",
-    hooks: [{
-      type: "command", 
-      command: scriptPath,
-      timeout: 180
-    }]
-  })
-} as const;
-
-// =============================================================================
-// CLI MESSAGES AND STATUS TEMPLATES
-// =============================================================================
-
-/**
- * Command-line interface messages
- */
-export const CLI_MESSAGES = {
-  INSTALLATION: {
-    STARTING: '🚀 Installing Claude Memory System with Chroma...',
-    SUCCESS: '🎉 Installation complete! Vector database ready.',
-    HOOKS_INSTALLED: '✅ Installed hooks to ~/.claude-mem/hooks/',
-    MCP_CONFIGURED: (path: string) => `✅ Configured MCP memory server in ${path}`,
-    EMBEDDED_READY: '🧠 Chroma initialized for persistent semantic memory',
-    ALREADY_INSTALLED: '⚠️  Claude Memory hooks are already installed.',
-    USE_FORCE: '   Use --force to overwrite existing installation.',
-    SETTINGS_WRITTEN: (type: string, path: string) => 
-      `✅ Installed hooks in ${type} settings\n   Settings file: ${path}`
-  },
-
-  NEXT_STEPS: [
-    '1. Restart Claude Code to load the new hooks',
-    '2. Use `/clear` and `/compact` in Claude Code to save and compress session memories', 
-    '3. New sessions will automatically load relevant context'
-  ],
-
-  ERRORS: {
-    HOOKS_NOT_FOUND: '❌ Hook source files not found',
-    SETTINGS_WRITE_FAILED: (path: string, error: string) => 
-      `❌ Failed to write settings file: ${error}\n   Path: ${path}`,
-    MCP_CONFIG_PARSE_FAILED: (error: string) => 
-      `⚠️  Warning: Could not parse existing MCP config: ${error}`,
-    MCP_CONFIG_WRITE_FAILED: (error: string) => 
-      `⚠️  Warning: Could not write MCP config: ${error}`,
-    COMPRESSION_FAILED: (error: string) => `❌ Compression failed: ${error}`,
-    CONTEXT_LOAD_FAILED: (error: string) => `❌ Failed to load context: ${error}`
-  },
-
-  STATUS: {
-    NO_INDEX: '📚 No memory index found. Starting fresh session.',
-    RECENT_MEMORIES: '🧠 Recent memories from previous sessions:',
-    MEMORY_COUNT: (count: number) => `📚 Showing ${count} most recent memories`,
-    FULL_CONTEXT_AVAILABLE: '💡 Full context available via MCP memory tools'
-  }
-} as const;
-
 // =============================================================================
 // DEBUG AND LOGGING TEMPLATES
 // =============================================================================
@@ -100,107 +17,9 @@ export const DEBUG_MESSAGES = {
  SESSION_ID: (id: string) => `🔍 Session ID: ${id}`,
  PROJECT_NAME: (name: string) => `📝 PROJECT NAME: ${name}`,
  CLAUDE_SDK_CALL: '🤖 Calling Claude SDK to analyze and populate memory database...',
-  TRANSCRIPT_STATS: (size: number, count: number) => 
+  TRANSCRIPT_STATS: (size: number, count: number) =>
    `📊 Transcript size: ${size} characters, ${count} messages`,
  COMPRESSION_COMPLETE: (count: number) => `✅ COMPRESSION COMPLETE\n  Total summaries extracted: ${count}`,
  CLAUDE_PATH_FOUND: (path: string) => `🎯 Found Claude Code at: ${path}`,
  MCP_CONFIG_USED: (path: string) => `📋 Using MCP config: ${path}`
 } as const;
-
-// =============================================================================
-// SEARCH AND QUERY TEMPLATES
-// =============================================================================
-
-/**
- * Memory database search templates
- */
-export const SEARCH_TEMPLATES = {
-  SEARCH_SCRIPT: (query: string) => `
-import { query } from "@anthropic-ai/claude-code";
-
-const searchQuery = process.env.SEARCH_QUERY || '';
-
-const result = await query({
-    prompt: "Search for: " + searchQuery,
-    options: {
-        mcpConfig: "~/.claude/.mcp.json",
-        allowedTools: ["mcp__claude-mem__chroma_query_documents"],
-        outputFormat: "json"
-    }
-});
-`,
-
-  SEARCH_PREFIX: "Search for: "
-} as const;
-
-// =============================================================================
-// CHROMA INTEGRATION CONSTANTS
-// =============================================================================
-
-/**
- * Chroma collection names for documents
- */
-export const CHROMA_COLLECTIONS = {
-  DOCUMENTS: 'claude_mem_documents',
-  MEMORIES: 'claude_mem_memories'
-} as const;
-
-/**
- * Default Chroma configuration values
- */
-export const CHROMA_DEFAULTS = {
-  HOST: 'localhost:8000',
-  COLLECTION: 'claude_mem_documents'
-} as const;
-
-/**
- * Chroma-specific CLI messages
- */
-export const CHROMA_MESSAGES = {
-  CONNECTION: {
-    CONNECTING: '🔗 Connecting to Chroma server...',
-    CONNECTED: '✅ Connected to Chroma successfully',
-    FAILED: (error: string) => `❌ Failed to connect to Chroma: ${error}`,
-    DISCONNECTED: '👋 Disconnected from Chroma'
-  },
-  
-  SEARCH: {
-    SEMANTIC_SEARCH: '🧠 Using semantic search with Chroma...',
-    KEYWORD_SEARCH: '🔍 Using keyword search with Chroma...',
-    HYBRID_SEARCH: '🔬 Using hybrid search with Chroma...',
-    RESULTS_FOUND: (count: number) => `📊 Found ${count} results in Chroma`
-  },
-  
-  SETUP: {
-    STARTING_CHROMA: '🚀 Starting Chroma instance...',
-    CHROMA_READY: '✅ Chroma is ready and accepting connections',
-    INITIALIZING_COLLECTIONS: '📋 Initializing document collections...'
-  }
-} as const;
-
-/**
- * Chroma error messages
- */
-export const CHROMA_ERRORS = {
-  CONNECTION_FAILED: 'Could not establish connection to Chroma server',
-  MCP_SERVER_NOT_FOUND: 'Chroma MCP server not found',
-  INVALID_COLLECTION: (collection: string) => `Invalid Chroma collection: ${collection}`,
-  QUERY_FAILED: (query: string, error: string) => `Query failed for '${query}': ${error}`,
-  DOCUMENT_CREATION_FAILED: (id: string) => `Failed to create document '${id}' in Chroma`,
-  COLLECTION_CREATION_FAILED: (name: string) => `Failed to create collection '${name}' in Chroma`
-} as const;
-
-/**
- * Export all core constants for easy importing
- */
-export const CONSTANTS = {
-  HOOK_CONFIG_TEMPLATES,
-  CLI_MESSAGES,
-  DEBUG_MESSAGES,
-  SEARCH_TEMPLATES,
-  // Chroma constants
-  CHROMA_COLLECTIONS,
-  CHROMA_DEFAULTS,
-  CHROMA_MESSAGES,
-  CHROMA_ERRORS
-} as const;
@@ -1,238 +0,0 @@
-/**
- * ChunkManager - Handles intelligent chunking of large transcripts
- * 
- * This class manages the splitting of large filtered transcripts into chunks
- * that fit within Claude's 32k token limit while preserving conversation context
- * and maintaining message integrity.
- */
-
-export interface ChunkMetadata {
-  chunkNumber: number;
-  totalChunks: number;
-  startIndex: number;
-  endIndex: number;
-  messageCount: number;
-  estimatedTokens: number;
-  sizeBytes: number;
-  hasOverlap: boolean;
-  overlapMessages?: number;
-  firstTimestamp?: string;
-  lastTimestamp?: string;
-}
-
-export interface ChunkingOptions {
-  maxTokensPerChunk?: number;  // default: 28000 (leaving 4k buffer)
-  maxBytesPerChunk?: number;   // default: 98000 (98KB)
-  preserveContext?: boolean;   // keep context overlap between chunks
-  contextOverlap?: number;     // messages to repeat (default: 2)
-  parallel?: boolean;          // process chunks in parallel
-}
-
-export interface ChunkedMessage {
-  content: string;
-  estimatedTokens: number;
-}
-
-export class ChunkManager {
-  private static readonly DEFAULT_MAX_TOKENS = 22400;  // Reduced by 20% from 28000
-  private static readonly DEFAULT_MAX_BYTES = 78400;   // Reduced by 20% from 98000
-  private static readonly DEFAULT_CONTEXT_OVERLAP = 2;
-  private static readonly CHARS_PER_TOKEN_ESTIMATE = 3.5;
-
-  private options: Required<ChunkingOptions>;
-
-  constructor(options: ChunkingOptions = {}) {
-    this.options = {
-      maxTokensPerChunk: options.maxTokensPerChunk ?? ChunkManager.DEFAULT_MAX_TOKENS,
-      maxBytesPerChunk: options.maxBytesPerChunk ?? ChunkManager.DEFAULT_MAX_BYTES,
-      preserveContext: options.preserveContext ?? true,
-      contextOverlap: options.contextOverlap ?? ChunkManager.DEFAULT_CONTEXT_OVERLAP,
-      parallel: options.parallel ?? false
-    };
-  }
-
-  /**
-   * Estimates token count for a given text
-   * Uses rough approximation of 3.5 characters per token
-   */
-  public estimateTokenCount(text: string): number {
-    return Math.ceil(text.length / ChunkManager.CHARS_PER_TOKEN_ESTIMATE);
-  }
-
-  /**
-   * Parses the filtered output format into structured messages
-   * Format: "- content"
-   */
-  public parseFilteredOutput(filteredContent: string): ChunkedMessage[] {
-    const lines = filteredContent.split('\n').filter(line => line.trim());
-    const messages: ChunkedMessage[] = [];
-
-    for (const line of lines) {
-      // Parse format: "- content"
-      if (line.startsWith('- ')) {
-        const content = line.substring(2); // Remove "- " prefix
-        messages.push({
-          content,
-          estimatedTokens: this.estimateTokenCount(content)
-        });
-      }
-    }
-
-    return messages;
-  }
-
-  /**
-   * Chunks the filtered transcript into manageable pieces
-   */
-  public chunkTranscript(filteredContent: string): Array<{ content: string; metadata: ChunkMetadata }> {
-    const messages = this.parseFilteredOutput(filteredContent);
-    const chunks: Array<{ content: string; metadata: ChunkMetadata }> = [];
-    
-    let currentChunk: ChunkedMessage[] = [];
-    let currentTokens = 0;
-    let currentBytes = 0;
-    let chunkStartIndex = 0;
-
-    for (let i = 0; i < messages.length; i++) {
-      const message = messages[i];
-      const messageText = this.formatMessage(message);
-      const messageBytes = Buffer.byteLength(messageText, 'utf8');
-      const messageTokens = message.estimatedTokens;
-
-      // Check if adding this message would exceed limits
-      if (currentChunk.length > 0 && 
-          (currentTokens + messageTokens > this.options.maxTokensPerChunk ||
-           currentBytes + messageBytes > this.options.maxBytesPerChunk)) {
-        
-        // Save current chunk
-        const chunkContent = this.formatChunk(currentChunk);
-        chunks.push({
-          content: chunkContent,
-          metadata: {
-            chunkNumber: chunks.length + 1,
-            totalChunks: 0, // Will be updated after all chunks are created
-            startIndex: chunkStartIndex,
-            endIndex: i - 1,
-            messageCount: currentChunk.length,
-            estimatedTokens: currentTokens,
-            sizeBytes: currentBytes,
-            hasOverlap: false
-          }
-        });
-
-        // Start new chunk with optional context overlap
-        currentChunk = [];
-        currentTokens = 0;
-        currentBytes = 0;
-        chunkStartIndex = i;
-
-        // Add overlap messages from previous chunk if enabled
-        if (this.options.preserveContext && chunks.length > 0) {
-          const overlapStart = Math.max(0, i - this.options.contextOverlap);
-          for (let j = overlapStart; j < i; j++) {
-            const overlapMessage = messages[j];
-            const overlapText = this.formatMessage(overlapMessage);
-            currentChunk.push(overlapMessage);
-            currentTokens += overlapMessage.estimatedTokens;
-            currentBytes += Buffer.byteLength(overlapText, 'utf8');
-          }
-          
-          if (currentChunk.length > 0) {
-            // Mark that this chunk has overlap
-            chunkStartIndex = overlapStart;
-          }
-        }
-      }
-
-      // Add message to current chunk
-      currentChunk.push(message);
-      currentTokens += messageTokens;
-      currentBytes += messageBytes;
-    }
-
-    // Save final chunk if it has content
-    if (currentChunk.length > 0) {
-      const chunkContent = this.formatChunk(currentChunk);
-      chunks.push({
-        content: chunkContent,
-        metadata: {
-          chunkNumber: chunks.length + 1,
-          totalChunks: 0,
-          startIndex: chunkStartIndex,
-          endIndex: messages.length - 1,
-          messageCount: currentChunk.length,
-          estimatedTokens: currentTokens,
-          sizeBytes: currentBytes,
-          hasOverlap: this.options.preserveContext && chunks.length > 0
-        }
-      });
-    }
-
-    // Update total chunks count in metadata
-    chunks.forEach(chunk => {
-      chunk.metadata.totalChunks = chunks.length;
-    });
-
-    return chunks;
-  }
-
-  /**
-   * Formats a single message back to the filtered output format
-   */
-  private formatMessage(message: ChunkedMessage): string {
-    return `- ${message.content}`;
-  }
-
-  /**
-   * Formats a chunk of messages
-   */
-  private formatChunk(messages: ChunkedMessage[]): string {
-    return messages.map(m => this.formatMessage(m)).join('\n');
-  }
-
-  /**
-   * Creates a header for a chunk file with metadata
-   */
-  public createChunkHeader(metadata: ChunkMetadata): string {
-    const lines = [];
-
-    // Add timestamp range if available, otherwise chunk number
-    if (metadata.firstTimestamp && metadata.lastTimestamp) {
-      lines.push(`# ${metadata.firstTimestamp} to ${metadata.lastTimestamp} (chunk ${metadata.chunkNumber}/${metadata.totalChunks})`);
-    } else {
-      lines.push(`# Chunk ${metadata.chunkNumber} of ${metadata.totalChunks}`);
-    }
-
-    return lines.join('\n') + '\n';
-  }
-
-  /**
-   * Checks if content needs chunking based on size
-   */
-  public needsChunking(content: string): boolean {
-    const estimatedTokens = this.estimateTokenCount(content);
-    const sizeBytes = Buffer.byteLength(content, 'utf8');
-    
-    return estimatedTokens > this.options.maxTokensPerChunk || 
-           sizeBytes > this.options.maxBytesPerChunk;
-  }
-
-  /**
-   * Gets chunking statistics for logging
-   */
-  public getChunkingStats(chunks: Array<{ metadata: ChunkMetadata }>): string {
-    const totalMessages = chunks.reduce((sum, c) => sum + c.metadata.messageCount, 0);
-    const totalTokens = chunks.reduce((sum, c) => sum + c.metadata.estimatedTokens, 0);
-    const totalBytes = chunks.reduce((sum, c) => sum + c.metadata.sizeBytes, 0);
-    
-    return [
-      `📊 Chunking Statistics:`,
-      `  • Total chunks: ${chunks.length}`,
-      `  • Total messages: ${totalMessages}`,
-      `  • Total estimated tokens: ${totalTokens.toLocaleString()}`,
-      `  • Total size: ${(totalBytes / 1024).toFixed(1)} KB`,
-      `  • Average tokens per chunk: ${Math.round(totalTokens / chunks.length).toLocaleString()}`,
-      `  • Average size per chunk: ${(totalBytes / chunks.length / 1024).toFixed(1)} KB`
-    ].join('\n');
-  }
-}
@@ -1,366 +0,0 @@
-/**
- * PromptOrchestrator - Single source of truth for all prompt generation
- * 
- * This class serves as the central orchestrator for generating different types of prompts
- * used throughout the claude-mem system. It provides clear, well-typed interfaces and
- * methods for creating prompts for LLM analysis, human context, and system integration.
- */
-
-import { createAnalysisPrompt } from '../../prompts/templates/analysis/AnalysisTemplates.js';
-
-// =============================================================================
-// CORE INTERFACES
-// =============================================================================
-
-/**
- * Context data for LLM analysis prompts
- */
-export interface AnalysisContext {
-  /** The transcript content to analyze */
-  transcriptContent: string;
-  /** Session identifier */
-  sessionId: string;
-  /** Project name for context */
-  projectName?: string;
-  /** Custom analysis instructions */
-  customInstructions?: string;
-  /** Compression trigger type */
-  trigger?: 'manual' | 'auto';
-  /** Original token count */
-  originalTokens?: number;
-  /** Target compression ratio */
-  targetCompressionRatio?: number;
-}
-
-/**
- * Context data for human-facing session prompts
- */
-export interface SessionContext {
-  /** Session identifier */
-  sessionId: string;
-  /** Source of the session start */
-  source: 'startup' | 'compact' | 'vscode' | 'web';
-  /** Project name */
-  projectName?: string;
-  /** Additional context to provide to the human */
-  additionalContext?: string;
-  /** Path to the transcript file */
-  transcriptPath?: string;
-  /** Working directory */
-  cwd?: string;
-}
-
-/**
- * Context data for hook response generation
- */
-export interface HookContext {
-  /** The hook event name */
-  hookEventName: string;
-  /** Session identifier */
-  sessionId: string;
-  /** Success status */
-  success: boolean;
-  /** Optional message */
-  message?: string;
-  /** Additional data specific to the hook */
-  data?: Record<string, unknown>;
-  /** Whether to continue processing */
-  shouldContinue?: boolean;
-  /** Reason for stopping if applicable */
-  stopReason?: string;
-}
-
-/**
- * Generated analysis prompt for LLM consumption
- */
-export interface AnalysisPrompt {
-  /** The formatted prompt text */
-  prompt: string;
-  /** Context used to generate the prompt */
-  context: AnalysisContext;
-  /** Prompt type identifier */
-  type: 'analysis';
-  /** Generated timestamp */
-  timestamp: string;
-}
-
-/**
- * Generated session prompt for human context
- */
-export interface SessionPrompt {
-  /** The formatted message text */
-  message: string;
-  /** Context used to generate the prompt */
-  context: SessionContext;
-  /** Prompt type identifier */
-  type: 'session';
-  /** Generated timestamp */
-  timestamp: string;
-}
-
-/**
- * Generated hook response
- */
-export interface HookResponse {
-  /** Whether to continue processing */
-  continue: boolean;
-  /** Reason for stopping if continue is false */
-  stopReason?: string;
-  /** Whether to suppress output */
-  suppressOutput?: boolean;
-  /** Hook-specific output data */
-  hookSpecificOutput?: Record<string, unknown>;
-  /** Context used to generate the response */
-  context: HookContext;
-  /** Response type identifier */
-  type: 'hook';
-  /** Generated timestamp */
-  timestamp: string;
-}
-
-// =============================================================================
-// PROMPT ORCHESTRATOR CLASS
-// =============================================================================
-
-/**
- * Central orchestrator for all prompt generation in the claude-mem system
- */
-export class PromptOrchestrator {
-  private projectName: string;
-
-  constructor(projectName = 'claude-mem') {
-    this.projectName = projectName;
-  }
-
-  /**
-   * Creates an analysis prompt for LLM processing of transcript content
-   */
-  public createAnalysisPrompt(context: AnalysisContext): AnalysisPrompt {
-    const timestamp = new Date().toISOString();
-    
-    const prompt = this.buildAnalysisPrompt(context);
-
-    return {
-      prompt,
-      context,
-      type: 'analysis',
-      timestamp,
-    };
-  }
-
-  /**
-   * Creates a session start prompt for human context
-   */
-  public createSessionStartPrompt(context: SessionContext): SessionPrompt {
-    const timestamp = new Date().toISOString();
-    
-    const message = this.buildSessionStartMessage(context);
-
-    return {
-      message,
-      context,
-      type: 'session',
-      timestamp,
-    };
-  }
-
-  /**
-   * Creates a hook response for system integration
-   */
-  public createHookResponse(context: HookContext): HookResponse {
-    const timestamp = new Date().toISOString();
-
-    const response = this.buildHookResponse(context);
-
-    return {
-      ...response,
-      context,
-      type: 'hook',
-      timestamp,
-    };
-  }
-
-  // =============================================================================
-  // PRIVATE PROMPT BUILDERS
-  // =============================================================================
-
-  private buildAnalysisPrompt(context: AnalysisContext): string {
-    const {
-      transcriptContent,
-      sessionId,
-      projectName = this.projectName,
-    } = context;
-
-    // Use project name as-is for consistency with directory names
-    const projectPrefix = projectName;
-
-    // Use the simple prompt with the transcript included
-    return createAnalysisPrompt(
-      transcriptContent,
-      sessionId,
-      projectPrefix
-    );
-  }
-
-  private buildSessionStartMessage(context: SessionContext): string {
-    const {
-      sessionId,
-      source,
-      projectName = this.projectName,
-      additionalContext,
-      transcriptPath,
-      cwd,
-    } = context;
-
-    let message = `## Session Started (${source})
-
-**Project**: ${projectName}  
-**Session ID**: ${sessionId}  `;
-
-    if (transcriptPath) {
-      message += `**Transcript**: ${transcriptPath}  `;
-    }
-
-    if (cwd) {
-      message += `**Working Directory**: ${cwd}  `;
-    }
-
-    if (additionalContext) {
-      message += `\n### Additional Context\n${additionalContext}`;
-    }
-
-    message += `\n\nMemory system is active and ready to preserve context across sessions.`;
-
-    return message;
-  }
-
-  private buildHookResponse(context: HookContext): Omit<HookResponse, 'context' | 'type' | 'timestamp'> {
-    const {
-      hookEventName,
-      success,
-      message,
-      data,
-      shouldContinue = success,
-      stopReason,
-    } = context;
-
-    const response: Omit<HookResponse, 'context' | 'type' | 'timestamp'> = {
-      continue: shouldContinue,
-      suppressOutput: false,
-    };
-
-    if (!shouldContinue && stopReason) {
-      response.stopReason = stopReason;
-    }
-
-    // Add hook-specific output based on event type
-    if (hookEventName === 'SessionStart') {
-      response.hookSpecificOutput = {
-        hookEventName: 'SessionStart',
-        additionalContext: message,
-        ...data,
-      };
-    } else if (data) {
-      response.hookSpecificOutput = data;
-    }
-
-    return response;
-  }
-
-  // =============================================================================
-  // UTILITY METHODS
-  // =============================================================================
-
-  /**
-   * Validates that an AnalysisContext has required fields
-   */
-  public validateAnalysisContext(context: Partial<AnalysisContext>): context is AnalysisContext {
-    return !!(context.transcriptContent && context.sessionId);
-  }
-
-  /**
-   * Validates that a SessionContext has required fields
-   */
-  public validateSessionContext(context: Partial<SessionContext>): context is SessionContext {
-    return !!(context.sessionId && context.source);
-  }
-
-  /**
-   * Validates that a HookContext has required fields
-   */
-  public validateHookContext(context: Partial<HookContext>): context is HookContext {
-    return !!(context.hookEventName && context.sessionId && typeof context.success === 'boolean');
-  }
-
-  /**
-   * Gets the project name for this orchestrator instance
-   */
-  public getProjectName(): string {
-    return this.projectName;
-  }
-
-  /**
-   * Sets a new project name for this orchestrator instance
-   */
-  public setProjectName(projectName: string): void {
-    this.projectName = projectName;
-  }
-}
-
-// =============================================================================
-// FACTORY FUNCTIONS
-// =============================================================================
-
-/**
- * Creates a new PromptOrchestrator instance
- */
-export function createPromptOrchestrator(projectName?: string): PromptOrchestrator {
-  return new PromptOrchestrator(projectName);
-}
-
-/**
- * Creates an analysis context from basic parameters
- */
-export function createAnalysisContext(
-  transcriptContent: string,
-  sessionId: string,
-  options: Partial<Omit<AnalysisContext, 'transcriptContent' | 'sessionId'>> = {}
-): AnalysisContext {
-  return {
-    transcriptContent,
-    sessionId,
-    ...options,
-  };
-}
-
-/**
- * Creates a session context from basic parameters
- */
-export function createSessionContext(
-  sessionId: string,
-  source: SessionContext['source'],
-  options: Partial<Omit<SessionContext, 'sessionId' | 'source'>> = {}
-): SessionContext {
-  return {
-    sessionId,
-    source,
-    ...options,
-  };
-}
-
-/**
- * Creates a hook context from basic parameters
- */
-export function createHookContext(
-  hookEventName: string,
-  sessionId: string,
-  success: boolean,
-  options: Partial<Omit<HookContext, 'hookEventName' | 'sessionId' | 'success'>> = {}
-): HookContext {
-  return {
-    hookEventName,
-    sessionId,
-    success,
-    ...options,
-  };
-}
@@ -1,128 +0,0 @@
-import { query } from '@anthropic-ai/claude-code';
-import fs from 'fs';
-import path from 'path';
-import os from 'os';
-import { getClaudePath } from '../../shared/settings.js';
-
-export interface TitleGenerationRequest {
-  sessionId: string;
-  projectName: string;
-  firstMessage: string;
-}
-
-export interface GeneratedTitle {
-  session_id: string;
-  generated_title: string;
-  timestamp: string;
-  project_name: string;
-}
-
-export class TitleGenerator {
-  private titlesIndexPath: string;
-
-  constructor() {
-    this.titlesIndexPath = path.join(os.homedir(), '.claude-mem', 'conversation-titles.jsonl');
-    this.ensureTitlesIndex();
-  }
-
-  private ensureTitlesIndex(): void {
-    const dir = path.dirname(this.titlesIndexPath);
-    if (!fs.existsSync(dir)) {
-      fs.mkdirSync(dir, { recursive: true });
-    }
-    if (!fs.existsSync(this.titlesIndexPath)) {
-      fs.writeFileSync(this.titlesIndexPath, '', 'utf-8');
-    }
-  }
-
-  async generateTitle(firstMessage: string): Promise<string> {
-    const prompt = `Generate a 3-7 word descriptive title for this conversation based on the first message.
-
-The title should:
- Capture the main topic or intent
- Be concise and descriptive
- Use proper capitalization
- Not include "Help with" or "Question about" prefixes
-
-First message: "${firstMessage.substring(0, 500)}"
-
-Respond with just the title, nothing else.`;
-
-    const response = await query({
-      prompt,
-      options: {
-        model: 'claude-3-5-haiku-20241022',
-        pathToClaudeCodeExecutable: getClaudePath(),
-      },
-    });
-
-    let title = '';
-    if (response && typeof response === 'object' && Symbol.asyncIterator in response) {
-      for await (const message of response) {
-        if (message?.content) title += message.content;
-        if (message?.text) title += message.text;
-      }
-    } else if (typeof response === 'string') {
-      title = response;
-    }
-
-    return title.trim().replace(/^["']|["']$/g, '');
-  }
-
-  async batchGenerateTitles(requests: TitleGenerationRequest[]): Promise<GeneratedTitle[]> {
-    const results: GeneratedTitle[] = [];
-    
-    for (const request of requests) {
-      try {
-        const title = await this.generateTitle(request.firstMessage);
-        
-        const generatedTitle: GeneratedTitle = {
-          session_id: request.sessionId,
-          generated_title: title,
-          timestamp: new Date().toISOString(),
-          project_name: request.projectName
-        };
-        
-        results.push(generatedTitle);
-        this.storeTitleInIndex(generatedTitle);
-      } catch (error) {
-        console.error(`Failed to generate title for ${request.sessionId}:`, error);
-      }
-    }
-    
-    return results;
-  }
-
-  private storeTitleInIndex(title: GeneratedTitle): void {
-    const line = JSON.stringify(title) + '\n';
-    fs.appendFileSync(this.titlesIndexPath, line, 'utf-8');
-  }
-
-  getExistingTitles(): Map<string, GeneratedTitle> {
-    const titles = new Map<string, GeneratedTitle>();
-    
-    if (!fs.existsSync(this.titlesIndexPath)) {
-      return titles;
-    }
-    
-    const content = fs.readFileSync(this.titlesIndexPath, 'utf-8');
-    const lines = content.trim().split('\n').filter(Boolean);
-    
-    for (const line of lines) {
-      try {
-        const title = JSON.parse(line) as GeneratedTitle;
-        titles.set(title.session_id, title);
-      } catch (error) {
-        // Skip invalid lines
-      }
-    }
-    
-    return titles;
-  }
-
-  getTitleForSession(sessionId: string): string | null {
-    const titles = this.getExistingTitles();
-    const title = titles.get(sessionId);
-    return title ? title.generated_title : null;
-  }
-}
@@ -0,0 +1,224 @@
+# Hook Prompts System
+
+This directory contains the centralized prompt configuration for all streaming hooks.
+
+## Quick Edit Guide
+
+**Want to change hook prompts?** Edit this file:
+```
+hook-prompts.config.ts
+```
+
+Then rebuild and reinstall:
+```bash
+bun run build
+bun run dev:install
+```
+
+## Files in This Directory
+
+### hook-prompts.config.ts
+**EDIT THIS FILE** to change prompt content.
+
+Contains:
+- `SYSTEM_PROMPT` - Initial instructions for SDK (190 lines)
+- `TOOL_MESSAGE` - Format for tool responses (10 lines)
+- `END_MESSAGE` - Session completion request (10 lines)
+- `HOOK_CONFIG` - Shared settings (truncation limits, SDK options)
+
+Uses `{{variableName}}` template syntax.
+
+### hook-prompt-renderer.ts
+**DON'T EDIT** unless adding new variables or changing rendering logic.
+
+Contains:
+- `renderSystemPrompt()` - Processes system prompt template
+- `renderToolMessage()` - Processes tool message template
+- `renderEndMessage()` - Processes end message template
+- Template substitution and auto-truncation logic
+
+### templates/context/ContextTemplates.ts
+Session start message formatting (separate from hook prompts).
+
+## Template Variables Reference
+
+### SYSTEM_PROMPT Variables
+```typescript
+{
+  project: string;        // Project name (e.g., "claude-mem-source")
+  sessionId: string;      // Claude Code session ID
+  date: string;           // YYYY-MM-DD format
+  userPrompt: string;     // Auto-truncated to 200 chars
+}
+```
+
+### TOOL_MESSAGE Variables
+```typescript
+{
+  toolName: string;       // Tool name (e.g., "Read", "Bash")
+  toolResponse: string;   // Auto-truncated to 20000 chars
+  userPrompt: string;     // Auto-truncated to 200 chars
+  timestamp: string;      // Full ISO timestamp
+  timeFormatted: string;  // HH:MM:SS format (auto-generated)
+}
+```
+
+### END_MESSAGE Variables
+```typescript
+{
+  project: string;        // Project name
+  sessionId: string;      // Claude Code session ID
+}
+```
+
+## Usage in Hooks
+
+### user-prompt-submit-streaming.js
+```javascript
+import { renderSystemPrompt, HOOK_CONFIG } from '../src/prompts/hook-prompt-renderer.js';
+
+const prompt = renderSystemPrompt({
+  project,
+  sessionId: session_id,
+  date,
+  userPrompt: prompt || ''
+});
+
+query({
+  prompt,
+  options: {
+    model: HOOK_CONFIG.sdk.model,
+    allowedTools: HOOK_CONFIG.sdk.allowedTools,
+    maxTokens: HOOK_CONFIG.sdk.maxTokensSystem
+  }
+});
+```
+
+### post-tool-use-streaming.js
+```javascript
+import { renderToolMessage, HOOK_CONFIG } from '../src/prompts/hook-prompt-renderer.js';
+
+const message = renderToolMessage({
+  toolName: tool_name,
+  toolResponse: toolResponseStr,
+  userPrompt: prompt || '',
+  timestamp: timestamp || new Date().toISOString()
+});
+
+query({
+  prompt: message,
+  options: {
+    model: HOOK_CONFIG.sdk.model,
+    maxTokens: HOOK_CONFIG.sdk.maxTokensTool
+  }
+});
+```
+
+### stop-streaming.js
+```javascript
+import { renderEndMessage, HOOK_CONFIG } from '../src/prompts/hook-prompt-renderer.js';
+
+const message = renderEndMessage({
+  project,
+  sessionId: claudeSessionId
+});
+
+query({
+  prompt: message,
+  options: {
+    model: HOOK_CONFIG.sdk.model,
+    maxTokens: HOOK_CONFIG.sdk.maxTokensEnd
+  }
+});
+```
+
+## Configuration Options
+
+Edit `HOOK_CONFIG` in `hook-prompts.config.ts`:
+
+```typescript
+export const HOOK_CONFIG = {
+  // Truncation limits for template variables
+  maxUserPromptLength: 200,      // Increase to show more context
+  maxToolResponseLength: 20000,  // Increase for larger outputs
+
+  // SDK configuration
+  sdk: {
+    model: 'claude-sonnet-4-5',  // Change model version
+    allowedTools: ['Bash'],      // Add more tools if needed
+    maxTokensSystem: 8192,       // Token limit for system prompt
+    maxTokensTool: 8192,         // Token limit for tool messages
+    maxTokensEnd: 2048,          // Token limit for end message
+  },
+};
+```
+
+## Example: Editing a Prompt
+
+### Before
+```typescript
+export const TOOL_MESSAGE = `# Tool Response {{timeFormatted}}
+
+Tool: {{toolName}}
+User Context: "{{userPrompt}}"
+
+\`\`\`
+{{toolResponse}}
+\`\`\`
+
+Analyze and store if meaningful.`;
+```
+
+### After
+```typescript
+export const TOOL_MESSAGE = `# Analysis Request {{timeFormatted}}
+
+Executed: {{toolName}}
+Context: "{{userPrompt}}"
+Priority: High
+
+Output:
+\`\`\`
+{{toolResponse}}
+\`\`\`
+
+IMPORTANT: Only store if this contains:
+- New code patterns or logic
+- Architecture decisions
+- Error messages with solutions
+- Configuration changes
+
+Skip trivial operations.`;
+```
+
+### Apply Changes
+```bash
+bun run build && bun run dev:install
+```
+
+## Benefits
+
+### DRY Compliance
+- **Before**: 3 files with 188 lines of hardcoded prompts
+- **After**: 1 config file with all prompts centralized
+
+### Maintainability
+- Change prompts without touching hook implementation
+- Type-safe template variables
+- Consistent formatting across all hooks
+- Version-controlled prompt history
+
+### Flexibility
+- Easy A/B testing of different instructions
+- Simple to adjust truncation limits
+- Quick model/token configuration changes
+- Template variables prevent copy-paste errors
+
+## Full Documentation
+
+See `/Users/alexnewman/Scripts/claude-mem-source/docs/HOOK_PROMPTS.md` for:
+- Detailed editing guide
+- Troubleshooting common issues
+- Adding new template variables
+- Advanced customization
+- Migration notes
@@ -1,191 +0,0 @@
-/**
- * Claude Memory System - Prompt-Related Constants and Templates
- * 
- * This file contains all prompts, instructions, and output templates
- * for the analysis and context priming system.
- */
-
-import * as HookTemplates from './templates/hooks/HookTemplates.js';
-
-// =============================================================================
-// ANALYSIS PROMPTS AND TEMPLATES
-// =============================================================================
-
-/**
- * Entity naming patterns for the knowledge graph
- */
-export const ENTITY_NAMING_PATTERNS = {
-  component: "Component_Name",
-  decision: "Decision_Name", 
-  pattern: "Pattern_Name",
-  tool: "Tool_Name",
-  fix: "Fix_Name",
-  workflow: "Workflow_Name"
-} as const;
-
-/**
- * Available entity types for classification
- */
-export const ENTITY_TYPES = {
-  component: "component", // UI components, modules, services
-  pattern: "pattern",     // Architectural or design patterns
-  workflow: "workflow",   // Processes, pipelines, sequences
-  integration: "integration", // APIs, external services, data sources
-  concept: "concept",     // Abstract ideas, methodologies, principles
-  decision: "decision",   // Design choices, trade-offs, solutions
-  tool: "tool",          // Utilities, libraries, development tools
-  fix: "fix"             // Bug fixes, patches, workarounds
-} as const;
-
-/**
- * Standard observation fields for entities
- */
-export const OBSERVATION_FIELDS = [
-  "Core purpose: [what it fundamentally does]",
-  "Brief description: [one-line summary for session-start display]", 
-  "Implementation: [key technical details, code patterns]",
-  "Dependencies: [what it requires or builds upon]",
-  "Usage context: [when/why it's used]",
-  "Performance characteristics: [speed, reliability, constraints]",
-  "Integration points: [how it connects to other systems]",
-  "Keywords: [searchable terms for this concept]",
-  "Decision rationale: [why this approach was chosen]",
-  "Next steps: [what needs to be done next with this component]",
-  "Files modified: [list of files changed]",
-  "Tools used: [development tools/commands used]"
-] as const;
-
-/**
- * Relationship types for creating meaningful entity connections
- */
-export const RELATIONSHIP_TYPES = [
-  "executes_via", "orchestrates_through", "validates_using",
-  "provides_auth_to", "manages_state_for", "processes_events_from",
-  "caches_data_from", "routes_requests_to", "transforms_data_for",
-  "extends", "enhances_performance_of", "builds_upon",
-  "fixes_issue_in", "replaces", "optimizes",
-  "triggers_tool", "receives_result_from"
-] as const;
-
-
-
-// =============================================================================
-// CONTEXT PRIMING TEMPLATES
-// =============================================================================
-
-/**
- * System message templates for context priming
- */
-export const CONTEXT_TEMPLATES = {
-  PRIMARY_CONTEXT: (projectName: string) => 
-    `Context primed for project: ${projectName}. Access memories with chroma_query_documents(["${projectName}*"]) or chroma_get_documents(["document_id"]).`,
-  
-  RECENT_SESSIONS: (sessionList: string) =>
-    `Recent sessions available: ${sessionList}`,
-    
-  AVAILABLE_ENTITIES: (type: string, entities: string[], hasMore: boolean, moreCount: number) =>
-    `Available ${type} entities: ${entities.join(', ')}${hasMore ? ` (+${moreCount} more)` : ''}`,
-
-  SESSION_START_HEADER: '🧠 Active Working Context from Previous Sessions:',
-  SESSION_START_SEPARATOR: '═'.repeat(70),
-  
-  RESUME_INSTRUCTIONS: `💡 TO RESUME: Load active components with chroma_get_documents(["<exact_document_ids>"])
-📊 TO EXPLORE: Search related work with chroma_query_documents(["<keywords>"])`
-} as const;
-
-// =============================================================================
-// SESSION START OUTPUT TEMPLATES  
-// =============================================================================
-
-/**
- * Session start formatting templates
- */
-export const SESSION_START_TEMPLATES = {
-  FOCUS_LINE: (focus: string) => `📌 CURRENT FOCUS: ${focus}`,
-  LAST_WORKED: (timeAgo: string, projectName: string) => `Last worked: ${timeAgo} | Project: ${projectName}`,
-  
-  SECTIONS: {
-    COMPONENTS: '🎯 ACTIVE COMPONENTS (load these for context):',
-    DECISIONS: '🔄 RECENT DECISIONS & PATTERNS:',
-    TOOLS: '🛠️ TOOLS & INFRASTRUCTURE:',
-    FIXES: '🐛 RECENT FIXES:',
-    ACTIONS: '⚡ NEXT ACTIONS:'
-  },
-  
-  ACTION_PREFIX: '□ ',
-  ENTITY_BULLET: '• '
-} as const;
-
-/**
- * Time formatting for "time ago" displays
- */
-export const TIME_FORMATS = {
-  JUST_NOW: 'just now',
-  HOURS_AGO: (hours: number) => `${hours} hour${hours > 1 ? 's' : ''} ago`,
-  DAYS_AGO: (days: number) => `${days} day${days > 1 ? 's' : ''} ago`,
-  RECENTLY: 'recently'
-} as const;
-
-// =============================================================================
-// HOOK RESPONSE TEMPLATES
-// =============================================================================
-
-/**
- * Standard hook response structures for Claude Code integration
- */
-export const HOOK_RESPONSES = {
-  SUCCESS: (hookEventName: string, message: string) => ({
-    hookSpecificOutput: {
-      hookEventName,
-      status: "success",
-      message
-    },
-    suppressOutput: true
-  }),
-
-  SKIPPED: (hookEventName: string, message: string) => ({
-    hookSpecificOutput: {
-      hookEventName,
-      status: "skipped", 
-      message
-    },
-    suppressOutput: true
-  }),
-
-  BLOCKED: (reason: string) => ({
-    decision: "block",
-    reason
-  }),
-
-  CONTINUE: (hookEventName: string, additionalContext?: string) => ({
-    continue: true,
-    ...(additionalContext && {
-      hookSpecificOutput: {
-        hookEventName,
-        additionalContext
-      }
-    })
-  }),
-
-  ERROR: (reason: string) => ({
-    decision: "block",
-    reason
-  })
-} as const;
-
-/**
- * Pre-defined hook messages
- */
-export const HOOK_MESSAGES = {
-  COMPRESSION_SUCCESS: "Memory compression completed successfully",
-  COMPRESSION_FAILED: (stderr: string) => `Compression failed: ${stderr}`,
-  CONTEXT_LOADED: "Project context loaded successfully",
-  CONTEXT_SKIPPED: "Continuing session - context loading skipped",
-  NO_TRANSCRIPT: "No transcript path provided",
-  HOOK_ERROR: (error: string) => `Hook error: ${error}`
-} as const;
-
-/**
- * Export hook templates for direct usage
- */
-export { HookTemplates };
@@ -0,0 +1,159 @@
+/**
+ * Hook Prompt Renderer
+ *
+ * Simple template rendering for hook prompts.
+ * Handles variable substitution and auto-truncation.
+ */
+
+import {
+  PROMPTS,
+  HOOK_CONFIG,
+  type SystemPromptVariables,
+  type ToolMessageVariables,
+  type EndMessageVariables,
+} from './hook-prompts.config.js';
+
+// =============================================================================
+// TEMPLATE RENDERING
+// =============================================================================
+
+/**
+ * Simple template variable substitution
+ * Replaces {{variableName}} with actual values
+ */
+function substituteVariables(
+  template: string,
+  variables: Record<string, string>
+): string {
+  let result = template;
+
+  for (const [key, value] of Object.entries(variables)) {
+    const placeholder = `{{${key}}}`;
+    // Replace all occurrences of this placeholder
+    result = result.split(placeholder).join(value);
+  }
+
+  return result;
+}
+
+/**
+ * Truncate text with ellipsis if it exceeds maxLength
+ */
+function truncate(text: string, maxLength: number): string {
+  if (text.length <= maxLength) return text;
+  return text.slice(0, maxLength) + (text.length > maxLength ? '...' : '');
+}
+
+/**
+ * Format timestamp for tool message header
+ * Extracts HH:MM:SS from ISO timestamp
+ */
+function formatTime(timestamp: string): string {
+  const timePart = timestamp.split('T')[1];
+  if (!timePart) return '';
+  return timePart.slice(0, 8); // HH:MM:SS
+}
+
+// =============================================================================
+// PUBLIC RENDERING FUNCTIONS
+// =============================================================================
+
+/**
+ * Render system prompt for SDK session initialization
+ */
+export function renderSystemPrompt(
+  variables: SystemPromptVariables
+): string {
+  // Auto-truncate userPrompt
+  const userPromptTruncated = truncate(
+    variables.userPrompt,
+    HOOK_CONFIG.maxUserPromptLength
+  );
+
+  return substituteVariables(PROMPTS.system, {
+    project: variables.project,
+    sessionId: variables.sessionId,
+    date: variables.date,
+    userPrompt: userPromptTruncated,
+  });
+}
+
+/**
+ * Render tool message for SDK processing
+ */
+export function renderToolMessage(
+  variables: ToolMessageVariables
+): string {
+  // Auto-truncate userPrompt and toolResponse
+  const userPromptTruncated = truncate(
+    variables.userPrompt,
+    HOOK_CONFIG.maxUserPromptLength
+  );
+
+  const toolResponseTruncated = truncate(
+    variables.toolResponse,
+    HOOK_CONFIG.maxToolResponseLength
+  );
+
+  // Format timestamp
+  const timeFormatted = formatTime(variables.timestamp);
+
+  return substituteVariables(PROMPTS.tool, {
+    toolName: variables.toolName,
+    toolResponse: toolResponseTruncated,
+    userPrompt: userPromptTruncated,
+    timestamp: variables.timestamp,
+    timeFormatted,
+  });
+}
+
+/**
+ * Render end message for session completion
+ */
+export function renderEndMessage(
+  variables: EndMessageVariables
+): string {
+  return substituteVariables(PROMPTS.end, {
+    project: variables.project,
+    sessionId: variables.sessionId,
+  });
+}
+
+// =============================================================================
+// GENERIC RENDERER (for convenience)
+// =============================================================================
+
+export type PromptType = 'system' | 'tool' | 'end';
+
+export type PromptVariables<T extends PromptType> = T extends 'system'
+  ? SystemPromptVariables
+  : T extends 'tool'
+    ? ToolMessageVariables
+    : T extends 'end'
+      ? EndMessageVariables
+      : never;
+
+/**
+ * Generic prompt renderer - dispatches to specific renderer based on type
+ */
+export function renderPrompt<T extends PromptType>(
+  type: T,
+  variables: PromptVariables<T>
+): string {
+  switch (type) {
+    case 'system':
+      return renderSystemPrompt(variables as SystemPromptVariables);
+    case 'tool':
+      return renderToolMessage(variables as ToolMessageVariables);
+    case 'end':
+      return renderEndMessage(variables as EndMessageVariables);
+    default:
+      throw new Error(`Unknown prompt type: ${type}`);
+  }
+}
+
+// =============================================================================
+// EXPORTS
+// =============================================================================
+
+export { HOOK_CONFIG, PROMPTS };
@@ -0,0 +1,306 @@
+/**
+ * Hook Prompts Configuration
+ *
+ * Centralized configuration for all streaming hook prompts.
+ * This is the SINGLE SOURCE OF TRUTH for hook prompt content.
+ *
+ * EDITING GUIDE:
+ * - Use {{variableName}} for template variables
+ * - Available variables are listed in each prompt's interface
+ * - All prompts are processed through renderPrompt() function
+ * - Changes here apply to all hooks automatically after rebuild
+ *
+ * LIFECYCLE FLOW:
+ * 1. user-prompt-submit: Initializes SDK session with systemPrompt
+ * 2. post-tool-use: Feeds tool responses using toolMessage (repeats N times)
+ * 3. stop: Ends session and requests overview using endMessage
+ */
+
+// =============================================================================
+// TYPE DEFINITIONS
+// =============================================================================
+
+export interface SystemPromptVariables {
+  project: string;
+  sessionId: string;
+  date: string;
+  userPrompt: string; // Auto-truncated to maxUserPromptLength
+}
+
+export interface ToolMessageVariables {
+  toolName: string;
+  toolResponse: string; // Auto-truncated to maxToolResponseLength
+  userPrompt: string; // Auto-truncated to maxUserPromptLength
+  timestamp: string; // Full ISO timestamp
+  timeFormatted: string; // HH:MM:SS format
+}
+
+export interface EndMessageVariables {
+  project: string;
+  sessionId: string;
+}
+
+// =============================================================================
+// SHARED CONFIGURATION
+// =============================================================================
+
+export const HOOK_CONFIG = {
+  // Truncation limits for template variables
+  maxUserPromptLength: 200,
+  maxToolResponseLength: 20000,
+
+  // SDK configuration (used by hooks)
+  sdk: {
+    model: 'claude-sonnet-4-5',
+    allowedTools: ['Bash'],
+    maxTokensSystem: 8192,
+    maxTokensTool: 8192,
+    maxTokensEnd: 2048,
+  },
+} as const;
+
+// =============================================================================
+// PHASE 1: SYSTEM PROMPT (user-prompt-submit-streaming.js)
+// =============================================================================
+
+/**
+ * System prompt that initializes the SDK session.
+ * Instructs the SDK on how to process tool responses and store memories.
+ *
+ * Variables:
+ * - {{project}}: Project name (from cwd basename)
+ * - {{sessionId}}: Claude Code session ID
+ * - {{date}}: Current date (YYYY-MM-DD)
+ * - {{userPrompt}}: User's initial prompt (truncated to 200 chars)
+ */
+export const SYSTEM_PROMPT = `You are a semantic memory compressor for claude-mem. You process tool responses from an active Claude Code session and store the important ones as searchable, hierarchical memories.
+
+# SESSION CONTEXT
+- Project: {{project}}
+- Session: {{sessionId}}
+- Date: {{date}}
+- User Request: "{{userPrompt}}"
+
+# YOUR JOB
+
+## FIRST: Generate Session Title
+
+IMMEDIATELY generate a title and subtitle for this session based on the user request.
+
+Use this bash command:
+\`\`\`bash
+claude-mem update-session-metadata \\
+  --project "{{project}}" \\
+  --session "{{sessionId}}" \\
+  --title "Short title (3-6 words)" \\
+  --subtitle "One sentence description (max 20 words)"
+\`\`\`
+
+Example for "Help me add dark mode to my app":
+- Title: "Dark Mode Implementation"
+- Subtitle: "Adding theme toggle and dark color scheme support to the application"
+
+## THEN: Process Tool Responses
+
+You will receive a stream of tool responses. For each one:
+
+1. ANALYZE: Does this contain information worth remembering?
+2. DECIDE: Should I store this or skip it?
+3. EXTRACT: What are the key semantic concepts?
+4. DECOMPOSE: Break into title + subtitle + atomic facts + narrative
+5. STORE: Use bash to save the hierarchical memory
+6. TRACK: Keep count of stored memories (001, 002, 003...)
+
+# WHAT TO STORE
+
+Store these:
+- File contents with logic, algorithms, or patterns
+- Search results revealing project structure
+- Build errors or test failures with context
+- Code revealing architecture or design decisions
+- Git diffs with significant changes
+- Command outputs showing system state
+
+Skip these:
+- Simple status checks (git status with no changes)
+- Trivial edits (one-line config changes)
+- Repeated operations
+- Binary data or noise
+- Anything without semantic value
+
+# HIERARCHICAL MEMORY FORMAT
+
+Each memory has FOUR components:
+
+## 1. TITLE (3-8 words)
+A scannable headline that captures the core action or topic.
+Examples:
+- "SDK Transcript Cleanup Implementation"
+- "Hook System Architecture Analysis"
+- "ChromaDB Migration Planning"
+
+## 2. SUBTITLE (max 24 words)
+A concise, memorable summary that captures the essence of the change.
+Examples:
+- "Automatic transcript cleanup after SDK session completion prevents memory conversations from appearing in UI history"
+- "Four lifecycle hooks coordinate session events: start, prompt submission, tool processing, and completion"
+- "Data migration from SQLite to ChromaDB enables semantic search across compressed conversation memories"
+
+Guidelines:
+- Clear and descriptive
+- Focus on the outcome or benefit
+- Use active voice when possible
+- Keep it professional and informative
+
+## 3. ATOMIC FACTS (3-7 facts, 50-150 chars each)
+Individual, searchable statements that can be vector-embedded separately.
+Each fact is ONE specific piece of information.
+
+Examples:
+- "stop-streaming.js: Auto-deletes SDK transcripts after completion"
+- "Path format: ~/.claude/projects/{sanitized-cwd}/{sessionId}.jsonl"
+- "Uses fs.unlink with graceful error handling for missing files"
+- "Checks two transcript path formats for backward compatibility"
+
+Guidelines:
+- Start with filename or component when relevant
+- Be specific: include paths, function names, actual values
+- Each fact stands alone (no pronouns like "it" or "this")
+- 50-150 characters target
+- Focus on searchable technical details
+
+## 4. NARRATIVE (512-1024 tokens, same as current format)
+The full contextual story for deep dives:
+
+"In the {{project}} project, [action taken]. [Technical details: files, functions, concepts]. [Why this matters]."
+
+This is the detailed explanation for when someone needs full context.
+
+# STORAGE COMMAND FORMAT
+
+Store using this EXACT bash command structure:
+\`\`\`bash
+claude-mem store-memory \\
+  --id "{{project}}_{{sessionId}}_{{date}}_001" \\
+  --title "Your Title Here" \\
+  --subtitle "Your concise subtitle here" \\
+  --facts '["Fact 1 here", "Fact 2 here", "Fact 3 here"]' \\
+  --concepts '["concept1", "concept2", "concept3"]' \\
+  --files '["path/to/file1.js", "path/to/file2.ts"]' \\
+  --project "{{project}}" \\
+  --session "{{sessionId}}" \\
+  --date "{{date}}"
+\`\`\`
+
+CRITICAL FORMATTING RULES:
+- Use single quotes around JSON arrays: --facts '["item1", "item2"]'
+- Use double quotes inside the JSON arrays: "item"
+- Use double quotes around simple string values: --title "Title"
+- Escape any quotes in the content properly
+- Sequential numbering: 001, 002, 003, etc.
+
+Concepts: 2-5 broad categories (e.g., "hooks", "storage", "async-processing")
+Files: Actual file paths touched (e.g., "hooks/stop-streaming.js")
+
+# EXAMPLE MEMORY
+
+Tool response shows: [Read file hooks/stop-streaming.js with 167 lines of code implementing SDK cleanup]
+
+Your storage command:
+\`\`\`bash
+claude-mem store-memory \\
+  --id "claude-mem_abc123_2025-10-01_001" \\
+  --title "SDK Transcript Auto-Cleanup" \\
+  --subtitle "Automatic deletion of SDK transcripts after completion prevents memory conversations from appearing in UI history" \\
+  --facts '["stop-streaming.js: Deletes SDK transcript after overview generation", "Path: ~/.claude/projects/{sanitized-cwd}/{sessionId}.jsonl", "Uses fs.unlink with error handling for missing files", "Prevents memory conversations from polluting Claude Code UI"]' \\
+  --concepts '["cleanup", "SDK-lifecycle", "UX", "file-management"]' \\
+  --files '["hooks/stop-streaming.js"]' \\
+  --project "claude-mem" \\
+  --session "abc123" \\
+  --date "2025-10-01"
+\`\`\`
+
+# STATE TRACKING
+
+CRITICAL: Keep track of your memory counter across all tool messages.
+- Start at 001
+- Increment for each stored memory
+- Never repeat numbers
+- Each session has separate numbering
+
+# SESSION END
+
+At the end (when I send "SESSION ENDING"), generate an overview using:
+\`\`\`bash
+claude-mem store-overview --project "{{project}}" --session "{{sessionId}}" --content "2-3 sentence overview"
+\`\`\`
+
+# IMPORTANT REMINDERS
+
+- You're processing a DIFFERENT Claude Code session (not your own)
+- Use Bash tool to call claude-mem commands
+- Keep subtitles clear and informative (max 24 words)
+- Each fact is ONE specific thing (not multiple ideas)
+- Be selective - quality over quantity
+- Always increment memory numbers
+- Facts should be searchable (specific file names, paths, functions)
+
+Ready for tool responses.`;
+
+// =============================================================================
+// PHASE 2: TOOL MESSAGE (post-tool-use-streaming.js)
+// =============================================================================
+
+/**
+ * Message format for each tool response sent to the SDK.
+ * The SDK analyzes this and decides whether to store a memory.
+ *
+ * Variables:
+ * - {{timeFormatted}}: Time portion of timestamp (HH:MM:SS)
+ * - {{toolName}}: Name of the tool that was used
+ * - {{userPrompt}}: User's original prompt (truncated to 200 chars)
+ * - {{toolResponse}}: Full tool response (truncated to 20000 chars)
+ */
+export const TOOL_MESSAGE = `# Tool Response {{timeFormatted}}
+
+Tool: {{toolName}}
+User Context: "{{userPrompt}}"
+
+\`\`\`
+{{toolResponse}}
+\`\`\`
+
+Analyze and store if meaningful.`;
+
+// =============================================================================
+// PHASE 3: END MESSAGE (stop-streaming.js)
+// =============================================================================
+
+/**
+ * Message sent to SDK when session ends.
+ * Requests the SDK to generate and store a session overview.
+ *
+ * Variables:
+ * - {{project}}: Project name
+ * - {{sessionId}}: Claude Code session ID
+ */
+export const END_MESSAGE = `# SESSION ENDING
+
+Review our entire conversation. Generate a concise 2-3 sentence overview of what was accomplished.
+
+Store it using Bash:
+\`\`\`bash
+claude-mem store-overview --project "{{project}}" --session "{{sessionId}}" --content "YOUR_OVERVIEW_HERE"
+\`\`\`
+
+Focus on: what was done, current state, key decisions, outcomes.`;
+
+// =============================================================================
+// EXPORTS
+// =============================================================================
+
+export const PROMPTS = {
+  system: SYSTEM_PROMPT,
+  tool: TOOL_MESSAGE,
+  end: END_MESSAGE,
+} as const;
@@ -1,30 +0,0 @@
-/**
- * Prompts Module - Single source of truth for all prompt generation
- * 
- * This module provides a centralized system for generating prompts across
- * the claude-mem system. It includes the core PromptOrchestrator class
- * and all related TypeScript interfaces.
- */
-
-// Export all interfaces
-export type {
-  AnalysisContext,
-  SessionContext,
-  HookContext,
-  AnalysisPrompt,
-  SessionPrompt,
-  HookResponse,
-} from '../core/orchestration/PromptOrchestrator.js';
-
-// Export the main class
-export {
-  PromptOrchestrator,
-} from '../core/orchestration/PromptOrchestrator.js';
-
-// Export factory functions
-export {
-  createPromptOrchestrator,
-  createAnalysisContext,
-  createSessionContext,
-  createHookContext,
-} from '../core/orchestration/PromptOrchestrator.js';
@@ -1,190 +0,0 @@
-# Claude Memory Templates
-
-This directory contains modular templates for the Claude Memory System, including LLM analysis prompts and system integration responses.
-
-## Files
-
-### AnalysisTemplates.ts
-The main template system for LLM analysis prompts. Contains clean, separated template functions for:
-
- **Entity extraction instructions** - Guidelines for identifying and categorizing technical entities
- **Relationship mapping instructions** - Rules for creating meaningful connections between entities
- **Output format specifications** - Exact format requirements for pipe-separated summaries
- **Example outputs** - Sample outputs to guide the LLM
- **MCP tool usage instructions** - Step-by-step MCP tool usage workflow
- **Dynamic content injection helpers** - Functions for injecting project/session context
-
-### HookTemplates.ts
-System integration templates for Claude Code hook responses. Provides standardized templates for:
-
- **Pre-compact hook responses** - Approve/block compression operations with proper formatting
- **Session-start hook responses** - Load and format context with rich memory information  
- **Pre-tool use hook responses** - Security policies and permission controls
- **Error handling templates** - User-friendly error messages with troubleshooting guidance
- **Progress indicators** - Status updates for long-running operations
- **Response validation** - Ensures compliance with Claude Code hook specifications
-
-### ContextTemplates.ts
-Human-readable formatting templates for user-facing messages during memory operations.
-
-### Legacy Templates
- `analysis-template.txt` - Legacy mustache-style template (deprecated)
- `session-start-template.txt` - Legacy mustache-style template (deprecated)
-
-## Architecture
-
-The new template system follows these principles:
-
-1. **Pure Functions** - Each template function takes context and returns formatted strings
-2. **Modular Design** - Complex prompts are broken into focused, reusable components
-3. **Type Safety** - Full TypeScript support with proper interfaces
-4. **Context Injection** - Dynamic content injection through helper functions
-5. **Composable Templates** - Build complex prompts by combining template sections
-
-## Usage
-
-### Hook Templates Usage
-```typescript
-import { 
-  createPreCompactSuccessResponse,
-  createSessionStartMemoryResponse,
-  createPreToolUseAllowResponse,
-  validateHookResponse 
-} from './HookTemplates.js';
-
-// Pre-compact hook: approve compression
-const preCompactResponse = createPreCompactSuccessResponse();
-console.log(JSON.stringify(preCompactResponse));
-// Output: {"continue": true, "suppressOutput": true}
-
-// Session start hook: load context with memories
-const sessionResponse = createSessionStartMemoryResponse({
-  projectName: 'claude-mem',
-  memoryCount: 15,
-  lastSessionTime: '2 hours ago',
-  recentComponents: ['HookTemplates', 'PromptOrchestrator'],
-  recentDecisions: ['Use TypeScript for type safety']
-});
-console.log(JSON.stringify(sessionResponse));
-
-// Pre-tool use: allow memory tools
-const toolResponse = createPreToolUseAllowResponse('Memory operations are always permitted');
-console.log(JSON.stringify(toolResponse));
-
-// Validate responses before sending
-const validation = validateHookResponse(preCompactResponse, 'PreCompact');
-if (!validation.isValid) {
-  console.error('Invalid response:', validation.errors);
-}
-```
-
-### Analysis Templates Usage
-```typescript
-import { buildCompleteAnalysisPrompt } from './AnalysisTemplates.js';
-
-const prompt = buildCompleteAnalysisPrompt(
-  'myproject',        // projectPrefix
-  'session123',       // sessionId
-  [],                // toolUseChains
-  '2024-01-01',      // timestamp (optional)
-  'archive.jsonl'    // archiveFilename (optional)
-);
-```
-
-### Individual Template Components
-```typescript
-import { 
-  createEntityExtractionInstructions,
-  createOutputFormatSpecification,
-  createExampleOutput
-} from './AnalysisTemplates.js';
-
-// Get just the entity extraction guidelines
-const entityInstructions = createEntityExtractionInstructions('myproject');
-
-// Get output format specification
-const outputFormat = createOutputFormatSpecification('2024-01-01', 'archive.jsonl');
-
-// Get example output
-const examples = createExampleOutput('myproject', 'session123');
-```
-
-### Context Injection
-```typescript
-import { 
-  injectProjectContext,
-  injectSessionContext,
-  validateTemplateContext 
-} from './AnalysisTemplates.js';
-
-// Validate context before using templates
-const context = { projectPrefix: 'myproject', sessionId: 'session123' };
-const errors = validateTemplateContext(context);
-if (errors.length > 0) {
-  console.error('Invalid context:', errors);
-}
-
-// Inject dynamic content into template strings
-let template = "Working on {{projectPrefix}} session {{sessionId}}";
-template = injectProjectContext(template, 'myproject');
-template = injectSessionContext(template, 'session123');
-```
-
-## Template Sections
-
-### Entity Extraction Instructions
- Categories of entities to extract (components, patterns, decisions, etc.)
- Naming conventions with project prefixes
- Entity type classifications
- Observation field templates
-
-### Relationship Mapping
- Available relationship types
- Active-voice relationship guidelines
- Graph connection strategies
-
-### Output Format
- Pipe-separated format specification
- Required fields and exact values
- Summary writing guidelines
-
-### MCP Tool Usage
- Step-by-step MCP tool workflow
- Entity creation instructions
- Relationship creation guidelines
-
-### Critical Requirements
- Entity count requirements (3-15 entities)
- Relationship count requirements (5-20 relationships)
- Output line requirements (3-10 summaries)
- Format validation rules
-
-## Benefits Over Legacy System
-
-1. **Maintainability** - Separated concerns make individual sections easy to update
-2. **Testability** - Pure functions can be unit tested independently
-3. **Reusability** - Template components can be reused across different contexts
-4. **Debugging** - Easy to isolate issues to specific template sections
-5. **Type Safety** - Full TypeScript support prevents runtime template errors
-6. **Performance** - No string parsing overhead, direct function composition
-
-## Migration from constants.ts
-
-The massive `createAnalysisPrompt` function in `constants.ts` has been refactored into this modular system:
-
-**Before** (130+ lines in single function):
-```typescript
-export function createAnalysisPrompt(...) {
-  // Massive template string with embedded logic
-  return `You are analyzing...${incrementalSection}${toolChains}...`;
-}
-```
-
-**After** (clean delegation):
-```typescript
-export function createAnalysisPrompt(...) {
-  return buildCompleteAnalysisPrompt(...);
-}
-```
-
-This maintains backward compatibility while providing a much cleaner, more maintainable internal structure.
@@ -1,118 +0,0 @@
-/**
- * Analysis Templates for LLM Instructions
- * 
- * Generates prompts for extracting memories from conversations and storing in Chroma
- */
-
-import Handlebars from 'handlebars';
-
-// =============================================================================
-// MAIN ANALYSIS PROMPT TEMPLATE
-// =============================================================================
-
-const ANALYSIS_PROMPT = `You are analyzing a Claude Code conversation transcript to create memories using the Chroma MCP memory system.
-
-YOUR TASK:
-1. Extract key learnings and accomplishments as natural language memories
-2. Store memories using mcp__claude-mem__chroma_add_documents
-3. Return a structured JSON response with the extracted summaries
-
-WHAT TO EXTRACT:
- Technical implementations (functions, classes, APIs, databases)
- Design patterns and architectural decisions
- Bug fixes and problem solutions
- Workflows, processes, and integrations
- Performance optimizations and improvements
-
-STORAGE INSTRUCTIONS:
-Call mcp__claude-mem__chroma_add_documents with:
- collection_name: "claude_memories"
- documents: Array of natural language descriptions
- ids: ["{{projectPrefix}}_{{sessionId}}_1", "{{projectPrefix}}_{{sessionId}}_2", ...]
- metadatas: Array with fields:
-  * type: component/pattern/workflow/integration/concept/decision/tool/fix
-  * keywords: Comma-separated search terms
-  * context: Brief situation description
-  * timestamp: "{{timestamp}}"
-  * session_id: "{{sessionId}}"
-
-ERROR HANDLING:
-If you get "IDs already exist" errors, use mcp__claude-mem__chroma_update_documents instead.
-If any tool calls fail, continue and return the JSON response anyway.
-
-Project: {{projectPrefix}}
-Session ID: {{sessionId}}
-
-Conversation to compress:`;
-
-// Compile template once
-const compiledAnalysisPrompt = Handlebars.compile(ANALYSIS_PROMPT, { noEscape: true });
-
-// =============================================================================
-// MAIN API FUNCTIONS
-// =============================================================================
-
-/**
- * Creates the comprehensive analysis prompt for memory extraction
- */
-export function buildComprehensiveAnalysisPrompt(
-  projectPrefix: string,
-  sessionId: string,
-  timestamp?: string,
-  archiveFilename?: string
-): string {
-  const context = {
-    projectPrefix,
-    sessionId,
-    timestamp: timestamp || new Date().toISOString(),
-    archiveFilename: archiveFilename || `${sessionId}.jsonl.archive`
-  };
-
-  return compiledAnalysisPrompt(context);
-}
-
-/**
- * Creates the analysis prompt
- */
-export function createAnalysisPrompt(
-  transcript: string,
-  sessionId: string,
-  projectPrefix: string,
-  timestamp?: string
-): string {
-  const prompt = buildComprehensiveAnalysisPrompt(
-    projectPrefix,
-    sessionId,
-    timestamp
-  );
-  
-  const responseFormat = `
-
-RESPONSE FORMAT:
-After storing memories in Chroma, return EXACTLY this JSON structure wrapped in tags:
-
-<JSONResponse>
-{
-  "overview": "2-3 sentence summary of session themes and accomplishments. Write for any developer to understand by organically defining jargon.",
-  "summaries": [
-    {
-      "text": "What was accomplished (start with action verb)",
-      "document_id": "${projectPrefix}_${sessionId}_1",
-      "keywords": "comma, separated, terms",
-      "timestamp": "${timestamp || new Date().toISOString()}",
-      "archive": "${sessionId}.jsonl.archive"
-    }
-  ]
-}
-</JSONResponse>
-
-IMPORTANT:
- Return 3-10 summaries based on conversation complexity
- Each summary should correspond to a memory you attempted to store
- If tool calls fail, still return the JSON response with summaries
- The JSON must be valid and complete
- Place NOTHING outside the <JSONResponse> tags
- Do not include any explanatory text before or after the JSON`;
-  
-  return prompt + '\n\n' + transcript + responseFormat;
-}
@@ -103,61 +103,6 @@ function makeLine(char: string = '─', width: number = getWrapWidth()): string
 // SESSION START MESSAGES
 // =============================================================================

-/**
- * Creates a welcoming session start message explaining what memories were loaded
- */
-export function createSessionStartMessage(
-  projectName: string,
-  memoryCount: number,
-  lastSessionTime?: string
-): string {
-  const width = getWrapWidth();
-  const timeInfo = lastSessionTime ? ` (last worked: ${lastSessionTime})` : '';
-
-  if (memoryCount === 0) {
-    return wrapText(
-      `🧠 Loading memories from previous sessions for ${projectName}${timeInfo}
-
-No relevant memories found - this appears to be your first session or a new project area.
-
-💡 Getting Started:
-• Start working and memories will be automatically created
-• At the end of your session, ask to compress and store the conversation
-• Next time you return, relevant context will be loaded automatically`,
-      width
-    );
-  }
-
-  const memoryText =
-    memoryCount === 1 ? 'relevant memory' : 'relevant memories';
-  return wrapText(
-    `🧠 Loading memories from previous sessions for ${projectName}${timeInfo}
-
-Found ${memoryCount} ${memoryText} to help continue your work.`,
-    width
-  );
-}
-
-// =============================================================================
-// OPERATION MESSAGES
-// =============================================================================
-
-/**
- * Creates a loading message during context retrieval
- */
-export function createLoadingMessage(operation: string): string {
-  const operations: Record<string, string> = {
-    searching: '🔍 Searching previous memories...',
-    loading: '📚 Loading relevant context...',
-    formatting: '✨ Organizing memories for display...',
-    compressing: '🗜️ Compressing session transcript...',
-    archiving: '📦 Archiving conversation...',
-  };
-
-  const width = getWrapWidth();
-  return wrapText(operations[operation] || `⏳ ${operation}...`, width);
-}
-
 /**
 * Creates a completion message after context operations
 */
@@ -263,28 +208,6 @@ export function formatTimeAgo(timestamp: string | Date): string {
  return date.toLocaleDateString();
 }

-/**
- * Creates summary text for memory operations
- */
-export function createOperationSummary(
-  operation: 'compress' | 'load' | 'search' | 'archive',
-  results: { count: number; duration?: number; details?: string }
-): string {
-  const { count, duration, details } = results;
-  const durationText = duration ? ` in ${duration}ms` : '';
-  const detailsText = details ? ` - ${details}` : '';
-
-  const templates = {
-    compress: `Compressed ${count} conversation turns${durationText}${detailsText}`,
-    load: `Loaded ${count} relevant memories${durationText}${detailsText}`,
-    search: `Found ${count} matching memories${durationText}${detailsText}`,
-    archive: `Archived ${count} conversation segments${durationText}${detailsText}`,
-  };
-
-  const width = getWrapWidth();
-  return wrapText(`📊 ${templates[operation]}`, width);
-}
-
 // =============================================================================
 // SESSION START TEMPLATE SYSTEM (data processing only)
 // =============================================================================
@@ -308,55 +231,6 @@ interface SessionGroup {
  memories: MemoryEntry[];
 }

-/**
- * Formats current date and time for session start
- */
-export function formatCurrentDateTime(): string {
-  const now = new Date();
-  const currentDateTime = now.toLocaleString('en-US', {
-    weekday: 'long',
-    year: 'numeric',
-    month: 'long',
-    day: 'numeric',
-    hour: '2-digit',
-    minute: '2-digit',
-    timeZoneName: 'short',
-  });
-
-  return `Current Date / Time: ${currentDateTime}\n`;
-}
-
-/**
- * Extracts overview section from JSON objects
- * Looks for objects with type "overview" and matching project
- */
-export function extractOverview(
-  recentObjects: any[],
-  projectName?: string
-): string | null {
-  // Find overview objects
-  const overviewObjects = recentObjects.filter(
-    (obj) => obj.type === 'overview'
-  );
-
-  if (overviewObjects.length === 0) {
-    return null;
-  }
-
-  // If project is specified, find overview for that project
-  if (projectName) {
-    const projectOverview = overviewObjects.find(
-      (obj) => obj.project === projectName
-    );
-    if (projectOverview) {
-      return projectOverview.content;
-    }
-  }
-
-  // Return the most recent overview if no project match
-  return overviewObjects[overviewObjects.length - 1]?.content || null;
-}
-
 /**
 * Interface for overview with timestamp
 */
@@ -377,66 +251,11 @@ interface SessionOverviewGroup {
  timeAgo?: string;
 }

-/**
- * Extracts multiple overviews with timestamps
- * Returns up to 'count' most recent overviews
- */
-export function extractOverviews(
-  recentObjects: any[],
-  count: number = 3,
-  projectName?: string
-): OverviewEntry[] {
-  // Find overview objects
-  const overviewObjects = recentObjects.filter(
-    (obj) => obj.type === 'overview'
-  );
-
-  if (overviewObjects.length === 0) {
-    return [];
-  }
-
-  // Filter by project if specified
-  let filteredOverviews = overviewObjects;
-  if (projectName) {
-    filteredOverviews = overviewObjects.filter(
-      (obj) => obj.project === projectName
-    );
-
-    // Fall back to all overviews if no project match
-    if (filteredOverviews.length === 0) {
-      filteredOverviews = overviewObjects;
-    }
-  }
-
-  // Take the last 'count' overviews
-  const recentOverviews = filteredOverviews.slice(-count);
-
-  // Process each overview with timestamp and session ID
-  return recentOverviews.map((obj) => {
-    const entry: OverviewEntry = {
-      content: obj.content || '',
-      sessionId: obj.sessionId || obj.session_id || 'unknown',
-    };
-
-    // Try to parse timestamp
-    const timestamp = parseTimestamp(obj);
-    if (timestamp) {
-      entry.timestamp = timestamp;
-      entry.timeAgo = formatRelativeTime(timestamp);
-    } else {
-      // Fallback if no timestamp
-      entry.timeAgo = 'Recently';
-    }
-
-    return entry;
-  }); // Show in original order (oldest to newest)
-}
-
 /**
 * Pure data processing function - converts JSON objects into structured memory entries
 * No formatting is done here, only data parsing and cleaning
 */
-export function processMemoryEntries(recentObjects: any[]): MemoryEntry[] {
+function processMemoryEntries(recentObjects: any[]): MemoryEntry[] {
  if (recentObjects.length === 0) {
    return [];
  }
@@ -1,185 +0,0 @@
-/**
- * Hook Templates Test
- * 
- * Basic validation tests for hook response templates to ensure they
- * generate valid responses that conform to Claude Code's hook system.
- */
-
-import {
-  createPreCompactSuccessResponse,
-  createPreCompactBlockedResponse,
-  createPreCompactApprovalResponse,
-  createSessionStartSuccessResponse,
-  createSessionStartEmptyResponse,
-  createSessionStartErrorResponse,
-  createSessionStartMemoryResponse,
-  createPreToolUseAllowResponse,
-  createPreToolUseDenyResponse,
-  createPreToolUseAskResponse,
-  createHookSuccessResponse,
-  createHookErrorResponse,
-  validateHookResponse,
-  createContextualHookResponse,
-  formatDuration,
-  createOperationSummary,
-  OPERATION_STATUS_TEMPLATES,
-  ERROR_RESPONSE_TEMPLATES
-} from './HookTemplates.js';
-
-// =============================================================================
-// PRE-COMPACT HOOK TESTS
-// =============================================================================
-
-console.log('Testing Pre-Compact Hook Templates...');
-
-// Test successful pre-compact response
-const preCompactSuccess = createPreCompactSuccessResponse();
-console.log('✓ Pre-compact success:', JSON.stringify(preCompactSuccess, null, 2));
-
-// Test blocked pre-compact response
-const preCompactBlocked = createPreCompactBlockedResponse('User requested to skip compression');
-console.log('✓ Pre-compact blocked:', JSON.stringify(preCompactBlocked, null, 2));
-
-// Test approval response
-const preCompactApproval = createPreCompactApprovalResponse('approve', 'Compression approved by policy');
-console.log('✓ Pre-compact approval:', JSON.stringify(preCompactApproval, null, 2));
-
-// =============================================================================
-// SESSION START HOOK TESTS
-// =============================================================================
-
-console.log('\nTesting Session Start Hook Templates...');
-
-// Test successful session start with context
-const sessionStartSuccess = createSessionStartSuccessResponse('Loaded 5 memories from previous sessions');
-console.log('✓ Session start success:', JSON.stringify(sessionStartSuccess, null, 2));
-
-// Test empty session start
-const sessionStartEmpty = createSessionStartEmptyResponse();
-console.log('✓ Session start empty:', JSON.stringify(sessionStartEmpty, null, 2));
-
-// Test error session start
-const sessionStartError = createSessionStartErrorResponse('Memory index corrupted');
-console.log('✓ Session start error:', JSON.stringify(sessionStartError, null, 2));
-
-// Test rich memory response
-const sessionStartMemory = createSessionStartMemoryResponse({
-  projectName: 'claude-mem',
-  memoryCount: 12,
-  lastSessionTime: '2 hours ago',
-  recentComponents: ['PromptOrchestrator', 'HookTemplates', 'MCPClient'],
-  recentDecisions: ['Use TypeScript for type safety', 'Implement embedded Weaviate']
-});
-console.log('✓ Session start memory:', JSON.stringify(sessionStartMemory, null, 2));
-
-// =============================================================================
-// PRE-TOOL USE HOOK TESTS
-// =============================================================================
-
-console.log('\nTesting Pre-Tool Use Hook Templates...');
-
-// Test allow response
-const preToolAllow = createPreToolUseAllowResponse('Tool execution approved by security policy');
-console.log('✓ Pre-tool allow:', JSON.stringify(preToolAllow, null, 2));
-
-// Test deny response
-const preToolDeny = createPreToolUseDenyResponse('Bash commands disabled in restricted mode');
-console.log('✓ Pre-tool deny:', JSON.stringify(preToolDeny, null, 2));
-
-// Test ask response
-const preToolAsk = createPreToolUseAskResponse('File operation requires user confirmation');
-console.log('✓ Pre-tool ask:', JSON.stringify(preToolAsk, null, 2));
-
-// =============================================================================
-// GENERIC HOOK TESTS
-// =============================================================================
-
-console.log('\nTesting Generic Hook Templates...');
-
-// Test basic success
-const genericSuccess = createHookSuccessResponse(false);
-console.log('✓ Generic success:', JSON.stringify(genericSuccess, null, 2));
-
-// Test basic error
-const genericError = createHookErrorResponse('Operation failed due to network timeout', true);
-console.log('✓ Generic error:', JSON.stringify(genericError, null, 2));
-
-// =============================================================================
-// VALIDATION TESTS
-// =============================================================================
-
-console.log('\nTesting Hook Response Validation...');
-
-// Test valid PreCompact response
-const preCompactValidation = validateHookResponse(preCompactSuccess, 'PreCompact');
-console.log('✓ PreCompact validation:', preCompactValidation);
-
-// Test invalid PreCompact response (with hookSpecificOutput)
-const invalidPreCompact = {
-  continue: true,
-  hookSpecificOutput: { hookEventName: 'PreCompact' }
-};
-const preCompactInvalidValidation = validateHookResponse(invalidPreCompact, 'PreCompact');
-console.log('✓ PreCompact invalid validation:', preCompactInvalidValidation);
-
-// Test valid SessionStart response
-const sessionStartValidation = validateHookResponse(sessionStartSuccess, 'SessionStart');
-console.log('✓ SessionStart validation:', sessionStartValidation);
-
-// =============================================================================
-// CONTEXTUAL HOOK RESPONSE TESTS
-// =============================================================================
-
-console.log('\nTesting Contextual Hook Responses...');
-
-// Test successful session start context
-const contextualSessionStart = createContextualHookResponse({
-  hookEventName: 'SessionStart',
-  sessionId: 'test-123',
-  success: true,
-  message: 'Successfully loaded 8 memories from previous claude-mem sessions'
-});
-console.log('✓ Contextual SessionStart:', JSON.stringify(contextualSessionStart, null, 2));
-
-// Test failed PreCompact context
-const contextualPreCompactFail = createContextualHookResponse({
-  hookEventName: 'PreCompact',
-  sessionId: 'test-123',
-  success: false,
-  message: 'Compression blocked: insufficient disk space'
-});
-console.log('✓ Contextual PreCompact fail:', JSON.stringify(contextualPreCompactFail, null, 2));
-
-// =============================================================================
-// UTILITY FUNCTION TESTS
-// =============================================================================
-
-console.log('\nTesting Utility Functions...');
-
-// Test duration formatting
-console.log('✓ Duration 500ms:', formatDuration(500));
-console.log('✓ Duration 5s:', formatDuration(5000));
-console.log('✓ Duration 90s:', formatDuration(90000));
-console.log('✓ Duration 2m30s:', formatDuration(150000));
-
-// Test operation summary
-console.log('✓ Operation summary success:', createOperationSummary('Memory compression', true, 5000, 15, 'entities extracted'));
-console.log('✓ Operation summary failure:', createOperationSummary('Context loading', false, 2000, 0, 'connection timeout'));
-
-// =============================================================================
-// TEMPLATE CONSTANT TESTS
-// =============================================================================
-
-console.log('\nTesting Template Constants...');
-
-// Test operation status templates
-console.log('✓ Compression complete:', OPERATION_STATUS_TEMPLATES.COMPRESSION_COMPLETE(25, 5000));
-console.log('✓ Context loaded:', OPERATION_STATUS_TEMPLATES.CONTEXT_LOADED(8));
-console.log('✓ Tool allowed:', OPERATION_STATUS_TEMPLATES.TOOL_ALLOWED('Bash'));
-
-// Test error response templates
-console.log('✓ File not found:', ERROR_RESPONSE_TEMPLATES.FILE_NOT_FOUND('/path/to/transcript.txt'));
-console.log('✓ Connection failed:', ERROR_RESPONSE_TEMPLATES.CONNECTION_FAILED('MCP memory server'));
-console.log('✓ Operation timeout:', ERROR_RESPONSE_TEMPLATES.OPERATION_TIMEOUT('compression', 30000));
-
-console.log('\n✅ All hook template tests completed successfully!');
@@ -1,546 +0,0 @@
-/**
- * Hook Templates for System Integration
- * 
- * This module provides standardized templates for hook responses that integrate
- * with Claude Code's hook system. These templates ensure consistent formatting
- * and proper JSON structure for different hook events.
- * 
- * Based on Claude Code Hook Documentation v2025
- */
-
-import { 
-  BaseHookResponse, 
-  PreCompactResponse, 
-  SessionStartResponse, 
-  PreToolUseResponse,
-  HookPayload,
-  PreCompactPayload,
-  SessionStartPayload
-} from '../../../shared/types.js';
-
-// =============================================================================
-// HOOK RESPONSE INTERFACES
-// =============================================================================
-
-/**
- * Context data for generating hook responses
- */
-export interface HookResponseContext {
-  /** The hook event name */
-  hookEventName: string;
-  /** Session identifier */
-  sessionId: string;
-  /** Whether the operation was successful */
-  success: boolean;
-  /** Optional message for the response */
-  message?: string;
-  /** Additional data specific to the hook type */
-  additionalData?: Record<string, unknown>;
-  /** Duration of the operation in milliseconds */
-  duration?: number;
-  /** Number of items processed */
-  itemCount?: number;
-}
-
-/**
- * Progress information for long-running operations
- */
-export interface OperationProgress {
-  /** Current step number */
-  current: number;
-  /** Total number of steps */
-  total: number;
-  /** Description of current step */
-  currentStep?: string;
-  /** Estimated time remaining in milliseconds */
-  estimatedRemaining?: number;
-}
-
-// =============================================================================
-// PRE-COMPACT HOOK TEMPLATES
-// =============================================================================
-
-/**
- * Creates a successful pre-compact response that allows compression to proceed
- * PreCompact hooks do NOT support hookSpecificOutput according to documentation
- */
-export function createPreCompactSuccessResponse(): PreCompactResponse {
-  return {
-    continue: true,
-    suppressOutput: true
-  };
-}
-
-/**
- * Creates a blocked pre-compact response that prevents compression
- */
-export function createPreCompactBlockedResponse(reason: string): PreCompactResponse {
-  return {
-    continue: false,
-    stopReason: reason,
-    suppressOutput: true
-  };
-}
-
-/**
- * Creates a pre-compact response with approval decision
- */
-export function createPreCompactApprovalResponse(
-  decision: 'approve' | 'block',
-  reason?: string
-): PreCompactResponse {
-  return {
-    decision,
-    reason,
-    continue: decision === 'approve',
-    suppressOutput: true
-  };
-}
-
-// =============================================================================
-// SESSION START HOOK TEMPLATES
-// =============================================================================
-
-/**
- * Creates a successful session start response with loaded context
- * SessionStart hooks DO support hookSpecificOutput
- */
-export function createSessionStartSuccessResponse(
-  additionalContext?: string
-): SessionStartResponse {
-  return {
-    continue: true,
-    suppressOutput: true,
-    hookSpecificOutput: {
-      hookEventName: 'SessionStart',
-      additionalContext
-    }
-  };
-}
-
-/**
- * Creates a session start response when no context is available
- */
-export function createSessionStartEmptyResponse(): SessionStartResponse {
-  return {
-    continue: true,
-    suppressOutput: true,
-    hookSpecificOutput: {
-      hookEventName: 'SessionStart',
-      additionalContext: 'Starting fresh session - no previous context available'
-    }
-  };
-}
-
-/**
- * Creates a session start response with error information
- */
-export function createSessionStartErrorResponse(error: string): SessionStartResponse {
-  return {
-    continue: true, // Continue even if context loading fails
-    suppressOutput: true,
-    hookSpecificOutput: {
-      hookEventName: 'SessionStart',
-      additionalContext: `Context loading encountered an issue: ${error}. Starting without previous context.`
-    }
-  };
-}
-
-/**
- * Creates a rich session start response with memory summary
- */
-export function createSessionStartMemoryResponse(memoryData: {
-  projectName: string;
-  memoryCount: number;
-  lastSessionTime?: string;
-  recentComponents?: string[];
-  recentDecisions?: string[];
-}): SessionStartResponse {
-  const { projectName, memoryCount, lastSessionTime, recentComponents = [], recentDecisions = [] } = memoryData;
-  
-  const timeInfo = lastSessionTime ? ` (last worked: ${lastSessionTime})` : '';
-  const contextParts: string[] = [];
-  
-  contextParts.push(`🧠 Loaded ${memoryCount} memories from previous sessions for ${projectName}${timeInfo}`);
-  
-  if (recentComponents.length > 0) {
-    contextParts.push(`\n🎯 Recent components: ${recentComponents.slice(0, 3).join(', ')}`);
-  }
-  
-  if (recentDecisions.length > 0) {
-    contextParts.push(`\n🔄 Recent decisions: ${recentDecisions.slice(0, 2).join(', ')}`);
-  }
-  
-  contextParts.push('\n💡 Use chroma_query_documents(["keywords"]) to find related work or chroma_get_documents(["document_id"]) to load specific content');
-  
-  return {
-    continue: true,
-    suppressOutput: true,
-    hookSpecificOutput: {
-      hookEventName: 'SessionStart',
-      additionalContext: contextParts.join('')
-    }
-  };
-}
-
-// =============================================================================
-// PRE-TOOL USE HOOK TEMPLATES
-// =============================================================================
-
-/**
- * Creates a pre-tool use response that allows the tool to execute
- */
-export function createPreToolUseAllowResponse(reason?: string): PreToolUseResponse {
-  return {
-    continue: true,
-    suppressOutput: true,
-    permissionDecision: 'allow',
-    permissionDecisionReason: reason
-  };
-}
-
-/**
- * Creates a pre-tool use response that blocks the tool execution
- */
-export function createPreToolUseDenyResponse(reason: string): PreToolUseResponse {
-  return {
-    continue: false,
-    stopReason: reason,
-    suppressOutput: true,
-    permissionDecision: 'deny',
-    permissionDecisionReason: reason
-  };
-}
-
-/**
- * Creates a pre-tool use response that asks for user confirmation
- */
-export function createPreToolUseAskResponse(reason: string): PreToolUseResponse {
-  return {
-    continue: true,
-    suppressOutput: false, // Show output so user can see the question
-    permissionDecision: 'ask',
-    permissionDecisionReason: reason
-  };
-}
-
-// =============================================================================
-// GENERIC HOOK RESPONSE TEMPLATES
-// =============================================================================
-
-/**
- * Creates a basic success response for any hook type
- */
-export function createHookSuccessResponse(suppressOutput = true): BaseHookResponse {
-  return {
-    continue: true,
-    suppressOutput
-  };
-}
-
-/**
- * Creates a basic error response for any hook type
- */
-export function createHookErrorResponse(
-  reason: string, 
-  suppressOutput = true
-): BaseHookResponse {
-  return {
-    continue: false,
-    stopReason: reason,
-    suppressOutput
-  };
-}
-
-/**
- * Creates a response with system message (warning/info for user)
- */
-export function createHookSystemMessageResponse(
-  message: string,
-  continueProcessing = true
-): BaseHookResponse & { systemMessage: string } {
-  return {
-    continue: continueProcessing,
-    suppressOutput: true,
-    systemMessage: message
-  };
-}
-
-// =============================================================================
-// OPERATION STATUS TEMPLATES
-// =============================================================================
-
-/**
- * Templates for different types of operation status messages
- */
-export const OPERATION_STATUS_TEMPLATES = {
-  // Compression operations
-  COMPRESSION_STARTED: 'Starting memory compression...',
-  COMPRESSION_ANALYZING: 'Analyzing transcript content...',
-  COMPRESSION_EXTRACTING: 'Extracting memories and connections...',
-  COMPRESSION_SAVING: 'Saving compressed memories...',
-  COMPRESSION_COMPLETE: (count: number, duration?: number) => 
-    `Memory compression complete. Extracted ${count} memories${duration ? ` in ${Math.round(duration/1000)}s` : ''}`,
-  
-  // Context loading operations
-  CONTEXT_LOADING: 'Loading previous session context...',
-  CONTEXT_SEARCHING: 'Searching for relevant memories...',
-  CONTEXT_FORMATTING: 'Organizing context for display...',
-  CONTEXT_LOADED: (count: number) => `Context loaded successfully. Found ${count} relevant memories`,
-  CONTEXT_EMPTY: 'No previous context found. Starting fresh session',
-  
-  // Tool operations
-  TOOL_CHECKING: (toolName: string) => `Checking permissions for ${toolName}...`,
-  TOOL_ALLOWED: (toolName: string) => `✅ ${toolName} execution approved`,
-  TOOL_BLOCKED: (toolName: string, reason: string) => `❌ ${toolName} blocked: ${reason}`,
-  
-  // General operations
-  OPERATION_STARTING: (operation: string) => `Starting ${operation}...`,
-  OPERATION_PROGRESS: (operation: string, current: number, total: number) => 
-    `${operation}: ${current}/${total} (${Math.round((current/total)*100)}%)`,
-  OPERATION_COMPLETE: (operation: string) => `✅ ${operation} completed successfully`,
-  OPERATION_FAILED: (operation: string, error: string) => `❌ ${operation} failed: ${error}`
-} as const;
-
-/**
- * Creates a progress message for long-running operations
- */
-export function createProgressMessage(
-  operation: string,
-  progress: OperationProgress
-): string {
-  const { current, total, currentStep, estimatedRemaining } = progress;
-  const percentage = Math.round((current / total) * 100);
-  
-  let message = `${operation}: ${current}/${total} (${percentage}%)`;
-  
-  if (currentStep) {
-    message += ` - ${currentStep}`;
-  }
-  
-  if (estimatedRemaining && estimatedRemaining > 1000) {
-    const seconds = Math.round(estimatedRemaining / 1000);
-    message += ` (${seconds}s remaining)`;
-  }
-  
-  return message;
-}
-
-// =============================================================================
-// ERROR RESPONSE TEMPLATES
-// =============================================================================
-
-/**
- * Standard error messages for different failure scenarios
- */
-export const ERROR_RESPONSE_TEMPLATES = {
-  // File system errors
-  FILE_NOT_FOUND: (path: string) => `File not found: ${path}`,
-  FILE_READ_ERROR: (path: string, error: string) => `Failed to read ${path}: ${error}`,
-  FILE_WRITE_ERROR: (path: string, error: string) => `Failed to write ${path}: ${error}`,
-  
-  // Network/connection errors
-  CONNECTION_FAILED: (service: string) => `Failed to connect to ${service}`,
-  CONNECTION_TIMEOUT: (service: string) => `Connection to ${service} timed out`,
-  
-  // Validation errors
-  INVALID_PAYLOAD: (field: string) => `Invalid or missing field: ${field}`,
-  INVALID_FORMAT: (expected: string, received: string) => `Expected ${expected}, received ${received}`,
-  
-  // Operation errors
-  OPERATION_TIMEOUT: (operation: string, timeout: number) => 
-    `${operation} timed out after ${timeout}ms`,
-  OPERATION_CANCELLED: (operation: string) => `${operation} was cancelled`,
-  INSUFFICIENT_PERMISSIONS: (operation: string) => 
-    `Insufficient permissions for ${operation}`,
-  
-  // Memory system errors
-  MEMORY_SYSTEM_UNAVAILABLE: 'Memory system is not available',
-  MEMORY_CORRUPTION: 'Memory index appears to be corrupted',
-  MEMORY_SEARCH_FAILED: (query: string) => `Memory search failed for query: "${query}"`,
-  
-  // Compression errors
-  COMPRESSION_FAILED: (stage: string) => `Compression failed during ${stage}`,
-  INVALID_TRANSCRIPT: 'Transcript file is invalid or corrupted',
-  
-  // General errors
-  UNKNOWN_ERROR: (context: string) => `An unexpected error occurred during ${context}`,
-  SYSTEM_ERROR: (error: string) => `System error: ${error}`
-} as const;
-
-/**
- * Creates a standardized error response with troubleshooting guidance
- */
-export function createDetailedErrorResponse(
-  operation: string,
-  error: string,
-  troubleshootingSteps: string[] = []
-): BaseHookResponse {
-  const baseMessage = `${operation} failed: ${error}`;
-  
-  const fullMessage = troubleshootingSteps.length > 0
-    ? `${baseMessage}\n\nTroubleshooting steps:\n${troubleshootingSteps.map(step => `• ${step}`).join('\n')}`
-    : baseMessage;
-  
-  return {
-    continue: false,
-    stopReason: fullMessage,
-    suppressOutput: false // Show error details to user
-  };
-}
-
-// =============================================================================
-// HOOK RESPONSE VALIDATION
-// =============================================================================
-
-/**
- * Validates that a hook response conforms to Claude Code expectations
- */
-export function validateHookResponse(
-  response: any,
-  hookType: string
-): { isValid: boolean; errors: string[] } {
-  const errors: string[] = [];
-  
-  // Check required fields
-  if (typeof response !== 'object' || response === null) {
-    errors.push('Response must be a valid JSON object');
-    return { isValid: false, errors };
-  }
-  
-  // Validate continue field
-  if (response.continue !== undefined && typeof response.continue !== 'boolean') {
-    errors.push('continue field must be a boolean');
-  }
-  
-  // Validate suppressOutput field
-  if (response.suppressOutput !== undefined && typeof response.suppressOutput !== 'boolean') {
-    errors.push('suppressOutput field must be a boolean');
-  }
-  
-  // Validate stopReason field
-  if (response.stopReason !== undefined && typeof response.stopReason !== 'string') {
-    errors.push('stopReason field must be a string');
-  }
-  
-  // Hook-specific validations
-  if (hookType === 'PreCompact') {
-    // PreCompact should not have hookSpecificOutput
-    if (response.hookSpecificOutput !== undefined) {
-      errors.push('PreCompact hooks do not support hookSpecificOutput');
-    }
-    
-    // Validate decision field if present
-    if (response.decision !== undefined && !['approve', 'block'].includes(response.decision)) {
-      errors.push('decision field must be "approve" or "block"');
-    }
-  }
-  
-  if (hookType === 'SessionStart') {
-    // Validate hookSpecificOutput structure
-    if (response.hookSpecificOutput) {
-      const hso = response.hookSpecificOutput;
-      if (hso.hookEventName !== 'SessionStart') {
-        errors.push('hookSpecificOutput.hookEventName must be "SessionStart"');
-      }
-      if (hso.additionalContext !== undefined && typeof hso.additionalContext !== 'string') {
-        errors.push('hookSpecificOutput.additionalContext must be a string');
-      }
-    }
-  }
-  
-  if (hookType === 'PreToolUse') {
-    // Validate permissionDecision field
-    if (response.permissionDecision !== undefined) {
-      if (!['allow', 'deny', 'ask'].includes(response.permissionDecision)) {
-        errors.push('permissionDecision must be "allow", "deny", or "ask"');
-      }
-    }
-  }
-  
-  return {
-    isValid: errors.length === 0,
-    errors
-  };
-}
-
-// =============================================================================
-// UTILITY FUNCTIONS
-// =============================================================================
-
-/**
- * Creates a hook response based on context and automatically handles hook-specific formatting
- */
-export function createContextualHookResponse(context: HookResponseContext): BaseHookResponse {
-  const { hookEventName, success, message, additionalData, duration, itemCount } = context;
-  
-  // Base response
-  const response: BaseHookResponse = {
-    continue: success,
-    suppressOutput: true
-  };
-  
-  // Add failure reason if not successful
-  if (!success && message) {
-    response.stopReason = message;
-    response.suppressOutput = false; // Show error to user
-  }
-  
-  // Handle hook-specific output
-  if (success && hookEventName === 'SessionStart' && message) {
-    return {
-      ...response,
-      hookSpecificOutput: {
-        hookEventName: 'SessionStart',
-        additionalContext: message
-      }
-    } as SessionStartResponse;
-  }
-  
-  // Handle PreCompact approval
-  if (hookEventName === 'PreCompact') {
-    return {
-      ...response,
-      decision: success ? 'approve' : 'block',
-      reason: message
-    } as PreCompactResponse;
-  }
-  
-  return response;
-}
-
-/**
- * Formats duration in milliseconds to human-readable format
- */
-export function formatDuration(milliseconds: number): string {
-  if (milliseconds < 1000) {
-    return `${milliseconds}ms`;
-  }
-  
-  const seconds = Math.round(milliseconds / 1000);
-  if (seconds < 60) {
-    return `${seconds}s`;
-  }
-  
-  const minutes = Math.floor(seconds / 60);
-  const remainingSeconds = seconds % 60;
-  return remainingSeconds > 0 ? `${minutes}m ${remainingSeconds}s` : `${minutes}m`;
-}
-
-/**
- * Creates a summary line for operation completion
- */
-export function createOperationSummary(
-  operation: string,
-  success: boolean,
-  duration?: number,
-  itemCount?: number,
-  details?: string
-): string {
-  const status = success ? '✅' : '❌';
-  const durationText = duration ? ` in ${formatDuration(duration)}` : '';
-  const itemText = itemCount ? ` (${itemCount} items)` : '';
-  const detailText = details ? ` - ${details}` : '';
-  
-  return `${status} ${operation}${itemText}${durationText}${detailText}`;
-}
@@ -1,364 +0,0 @@
-import * as p from '@clack/prompts';
-import { TranscriptParser } from './transcript-parser.js';
-import path from 'path';
-import fs from 'fs';
-
-/**
- * Conversation item for selection UI
- */
-export interface ConversationItem {
-  filePath: string;
-  sessionId: string;
-  timestamp: string;
-  messageCount: number;
-  gitBranch?: string;
-  cwd: string;
-  fileSize: number;
-  displayName: string;
-  projectName: string;
-  parsedDate: Date;
-  relativeDate: string;
-  dateGroup: string;
-}
-
-
-/**
- * Selection result
- */
-export interface SelectionResult {
-  selectedFiles: string[];
-  cancelled: boolean;
-}
-
-/**
- * Interactive conversation selector service
- */
-export class ConversationSelector {
-  private parser: TranscriptParser;
-  
-  constructor() {
-    this.parser = new TranscriptParser();
-  }
-  
-  /**
-   * Show interactive selection UI for conversations with improved flow
-   */
-  async selectConversations(): Promise<SelectionResult> {
-    p.intro('🧠 Claude History Import');
-    
-    const s = p.spinner();
-    s.start('Scanning for conversation files...');
-    
-    const conversationFiles = await this.parser.scanConversationFiles();
-    
-    if (conversationFiles.length === 0) {
-      s.stop('❌ No conversation files found');
-      p.outro('No conversation files found in Claude projects directory');
-      return { selectedFiles: [], cancelled: true };
-    }
-    
-    // Get metadata for each file
-    const conversations: ConversationItem[] = [];
-    for (const filePath of conversationFiles) {
-      try {
-        const metadata = await this.parser.getConversationMetadata(filePath);
-        const projectName = this.extractProjectName(filePath);
-        const parsedDate = this.parseTimestamp(metadata.timestamp, filePath);
-        const relativeDate = this.formatRelativeDate(parsedDate);
-        const dateGroup = this.getDateGroup(parsedDate);
-        
-        conversations.push({
-          filePath,
-          ...metadata,
-          projectName,
-          parsedDate,
-          relativeDate,
-          dateGroup,
-          displayName: this.createDisplayName(filePath, metadata)
-        });
-      } catch (e) {
-        // Skip invalid files silently
-      }
-    }
-    
-    if (conversations.length === 0) {
-      s.stop('❌ No valid conversation files found');
-      p.outro('No valid conversation files found');
-      return { selectedFiles: [], cancelled: true };
-    }
-    
-    s.stop(`Found ${conversations.length} conversation files`);
-    
-    // Sort by timestamp (newest first)
-    conversations.sort((a, b) => b.parsedDate.getTime() - a.parsedDate.getTime());
-    
-    // If there are too many conversations, offer filtering options first
-    let filteredConversations = conversations;
-    if (conversations.length > 100) {
-      const filterChoice = await p.select({
-        message: `Found ${conversations.length} conversations. How would you like to proceed?`,
-        options: [
-          { value: 'recent', label: 'Show recent (last 50)', hint: 'Most recent conversations' },
-          { value: 'project', label: 'Filter by project', hint: 'Select specific project first' },
-          { value: 'all', label: 'Show all', hint: `Display all ${conversations.length} conversations` }
-        ]
-      });
-      
-      if (p.isCancel(filterChoice)) {
-        p.cancel('Selection cancelled');
-        return { selectedFiles: [], cancelled: true };
-      }
-      
-      if (filterChoice === 'recent') {
-        filteredConversations = conversations.slice(0, 50);
-      } else if (filterChoice === 'project') {
-        const projectNames = [...new Set(conversations.map(c => c.projectName))].sort();
-        const selectedProject = await p.select({
-          message: 'Select project:',
-          options: projectNames.map(project => {
-            const count = conversations.filter(c => c.projectName === project).length;
-            return {
-              value: project,
-              label: project,
-              hint: `${count} conversation${count === 1 ? '' : 's'}`
-            };
-          })
-        });
-        
-        if (p.isCancel(selectedProject)) {
-          p.cancel('Selection cancelled');
-          return { selectedFiles: [], cancelled: true };
-        }
-        
-        filteredConversations = conversations.filter(c => c.projectName === selectedProject);
-      }
-    }
-    
-    // Conversation selection
-    const selectedConversations = await this.selectConversationsFromList(filteredConversations);
-    if (!selectedConversations || selectedConversations.length === 0) {
-      p.cancel('No conversations selected');
-      return { selectedFiles: [], cancelled: true };
-    }
-    
-    // Confirmation
-    const confirmed = await this.confirmSelection(selectedConversations);
-    if (!confirmed) {
-      p.cancel('Import cancelled');
-      return { selectedFiles: [], cancelled: true };
-    }
-    
-    p.outro(`Ready to import ${selectedConversations.length} conversations`);
-    return { selectedFiles: selectedConversations.map(c => c.filePath), cancelled: false };
-  }
-  
-  /**
-   * Extract project name from file path
-   */
-  private extractProjectName(filePath: string): string {
-    return path.basename(path.dirname(filePath));
-  }
-  
-  /**
-   * Safely parse timestamp with fallback to file modification time
-   */
-  private parseTimestamp(timestamp: string | undefined, filePath: string): Date {
-    // Try parsing the provided timestamp
-    if (timestamp) {
-      const date = new Date(timestamp);
-      if (!isNaN(date.getTime())) {
-        return date;
-      }
-    }
-    
-    // Fallback to file modification time
-    try {
-      const stats = fs.statSync(filePath);
-      return stats.mtime;
-    } catch (e) {
-      // Last resort: current time
-      return new Date();
-    }
-  }
-  
-  /**
-   * Format date as relative time (e.g., "2 days ago", "3 weeks ago")
-   */
-  private formatRelativeDate(date: Date): string {
-    const now = new Date();
-    const diffMs = now.getTime() - date.getTime();
-    const diffMinutes = Math.floor(diffMs / (1000 * 60));
-    const diffHours = Math.floor(diffMs / (1000 * 60 * 60));
-    const diffDays = Math.floor(diffMs / (1000 * 60 * 60 * 24));
-    const diffWeeks = Math.floor(diffDays / 7);
-    const diffMonths = Math.floor(diffDays / 30);
-    
-    if (diffMinutes < 1) return 'just now';
-    if (diffMinutes < 60) return `${diffMinutes}m ago`;
-    if (diffHours < 24) return `${diffHours}h ago`;
-    if (diffDays < 7) return `${diffDays}d ago`;
-    if (diffWeeks < 4) return `${diffWeeks}w ago`;
-    if (diffMonths < 12) return `${diffMonths}mo ago`;
-    
-    const diffYears = Math.floor(diffMonths / 12);
-    return `${diffYears}y ago`;
-  }
-  
-  /**
-   * Get date group for grouping conversations
-   */
-  private getDateGroup(date: Date): string {
-    const now = new Date();
-    const today = new Date(now.getFullYear(), now.getMonth(), now.getDate());
-    const yesterday = new Date(today.getTime() - 24 * 60 * 60 * 1000);
-    const thisWeekStart = new Date(today.getTime() - today.getDay() * 24 * 60 * 60 * 1000);
-    const lastWeekStart = new Date(thisWeekStart.getTime() - 7 * 24 * 60 * 60 * 1000);
-    const thisMonthStart = new Date(now.getFullYear(), now.getMonth(), 1);
-    
-    const conversationDate = new Date(date.getFullYear(), date.getMonth(), date.getDate());
-    
-    if (conversationDate.getTime() >= today.getTime()) {
-      return 'Today';
-    } else if (conversationDate.getTime() >= yesterday.getTime()) {
-      return 'Yesterday';
-    } else if (conversationDate.getTime() >= thisWeekStart.getTime()) {
-      return 'This Week';
-    } else if (conversationDate.getTime() >= lastWeekStart.getTime()) {
-      return 'Last Week';
-    } else if (conversationDate.getTime() >= thisMonthStart.getTime()) {
-      return 'This Month';
-    } else {
-      return 'Older';
-    }
-  }
-  
-  /**
-   * Create display name for conversation
-   */
-  private createDisplayName(filePath: string, metadata: any): string {
-    const parsedDate = this.parseTimestamp(metadata.timestamp, filePath);
-    const relativeDate = this.formatRelativeDate(parsedDate);
-    const sizeKB = Math.round(metadata.fileSize / 1024);
-    const branchInfo = metadata.gitBranch ? `${metadata.gitBranch}` : '';
-    
-    return `${relativeDate} • ${metadata.messageCount} msgs • ${sizeKB}KB${branchInfo ? ` • ${branchInfo}` : ''}`;
-  }
-  
-  
-  /**
-   * Select specific conversations from list
-   */
-  private async selectConversationsFromList(
-    conversations: ConversationItem[]
-  ): Promise<ConversationItem[] | null> {
-    // Group conversations by date for better organization
-    const groupedConversations = this.groupConversationsByDate(conversations);
-    const options = this.createGroupedOptions(groupedConversations, conversations);
-    
-    // Multi-select with select all/none shortcuts
-    const selectedIndices = await p.multiselect({
-      message: `Select conversations to import (${conversations.length} available, Space=toggle, Enter=confirm):`,
-      options,
-      required: false
-    });
-    
-    if (p.isCancel(selectedIndices)) return null;
-    
-    // Return selected conversations
-    const selected = selectedIndices as number[];
-    if (selected.length === 0) {
-      return [];
-    }
-    
-    return selected.map(i => conversations[i]);
-  }
-  
-  /**
-   * Confirm selection before processing
-   */
-  private async confirmSelection(conversations: ConversationItem[]): Promise<boolean> {
-    const totalSize = conversations.reduce((sum, c) => sum + c.fileSize, 0);
-    const sizeKB = Math.round(totalSize / 1024);
-    const projects = [...new Set(conversations.map(c => c.projectName))];
-    
-    const details = [
-      `${conversations.length} conversation${conversations.length === 1 ? '' : 's'}`,
-      `${projects.length} project${projects.length === 1 ? '' : 's'}: ${projects.join(', ')}`,
-      `Total size: ${sizeKB}KB`
-    ].join('\n');
-    
-    const confirmed = await p.confirm({
-      message: `Ready to import:\n\n${details}\n\nContinue?`,
-      initialValue: true
-    });
-    
-    return !p.isCancel(confirmed) && confirmed;
-  }
-  
-  /**
-   * Group conversations by date sections
-   */
-  private groupConversationsByDate(conversations: ConversationItem[]): Map<string, ConversationItem[]> {
-    const groups = new Map<string, ConversationItem[]>();
-    
-    for (const conv of conversations) {
-      const group = conv.dateGroup;
-      if (!groups.has(group)) {
-        groups.set(group, []);
-      }
-      groups.get(group)!.push(conv);
-    }
-    
-    return groups;
-  }
-  
-  /**
-   * Create options with date group headers
-   */
-  private createGroupedOptions(groupedConversations: Map<string, ConversationItem[]>, allConversations: ConversationItem[]) {
-    const options: any[] = [];
-    
-    // Add hint at top about selecting all/none
-    options.push({ 
-      value: 'hint', 
-      label: '💡 Use Space to toggle, A to select all, I to invert', 
-      disabled: true 
-    });
-    options.push({ value: 'separator-hint', label: '─'.repeat(60), disabled: true });
-    
-    // Define order of groups
-    const groupOrder = ['Today', 'Yesterday', 'This Week', 'Last Week', 'This Month', 'Older'];
-    
-    for (const groupName of groupOrder) {
-      const conversations = groupedConversations.get(groupName);
-      if (!conversations || conversations.length === 0) continue;
-      
-      // Add group header (disabled option for visual separation)
-      if (options.length > 2) { // Account for hint and separator
-        options.push({ value: `separator-${groupName}`, label: '─'.repeat(50), disabled: true });
-      }
-      options.push({ 
-        value: `header-${groupName}`, 
-        label: `${groupName} (${conversations.length})`, 
-        disabled: true 
-      });
-      
-      // Add conversations in this group
-      for (const conv of conversations) {
-        const index = allConversations.indexOf(conv);
-        const projectInfo = conv.projectName ? `[${conv.projectName}]` : '';
-        const workingDir = conv.cwd && conv.cwd !== 'undefined' ? path.basename(conv.cwd) : '';
-        const hint = `${projectInfo} ${workingDir}`.trim() || (conv.gitBranch ? `Branch: ${conv.gitBranch}` : '');
-        
-        options.push({
-          value: index,
-          label: `  ${conv.displayName}`,
-          hint: hint
-        });
-      }
-    }
-    
-    return options;
-  }
-  
-}
@@ -181,7 +181,9 @@ export class PathDiscovery {
      const packageJsonPath = require.resolve('claude-mem/package.json');
      this._packageRoot = dirname(packageJsonPath);
      return this._packageRoot;
-    } catch {}
+    } catch {
+      // Continue to next method
+    }

    // Method 2: Walk up from current module location
    const currentFile = fileURLToPath(import.meta.url);
@@ -190,15 +192,13 @@ export class PathDiscovery {
    for (let i = 0; i < 10; i++) {
      const packageJsonPath = join(currentDir, 'package.json');
      if (existsSync(packageJsonPath)) {
-        try {
-          const packageJson = require(packageJsonPath);
-          if (packageJson.name === 'claude-mem') {
-            this._packageRoot = currentDir;
-            return this._packageRoot;
-          }
-        } catch {}
+        const packageJson = require(packageJsonPath);
+        if (packageJson.name === 'claude-mem') {
+          this._packageRoot = currentDir;
+          return this._packageRoot;
+        }
      }
-      
+
      const parentDir = dirname(currentDir);
      if (parentDir === currentDir) break;
      currentDir = parentDir;
@@ -206,36 +206,46 @@ export class PathDiscovery {

    // Method 3: Try npm list command
    try {
-      const npmOutput = execSync('npm list -g claude-mem --json 2>/dev/null || npm list claude-mem --json 2>/dev/null', { 
-        encoding: 'utf8' 
+      const npmOutput = execSync('npm list -g claude-mem --json 2>/dev/null || npm list claude-mem --json 2>/dev/null', {
+        encoding: 'utf8'
      });
      const npmData = JSON.parse(npmOutput);
-      
+
      if (npmData.dependencies?.['claude-mem']?.resolved) {
        this._packageRoot = dirname(npmData.dependencies['claude-mem'].resolved);
        return this._packageRoot;
      }
-    } catch {}
+    } catch {
+      // Continue to error
+    }

    throw new Error('Cannot locate claude-mem package root. Ensure claude-mem is properly installed.');
  }

  /**
-   * Find hooks directory in the installed package
+   * Find hook templates directory in the installed package
+   *
+   * This returns the SOURCE templates directory that gets copied during installation
+   * to the runtime hooks directory (~/.claude-mem/hooks/)
   */
-  findPackageHooksDirectory(): string {
+  findPackageHookTemplatesDirectory(): string {
    const packageRoot = this.getPackageRoot();
-    const hooksDir = join(packageRoot, 'hooks');
-    
-    // Verify it contains expected hook files
-    const requiredHooks = ['pre-compact.js', 'session-start.js'];
-    for (const hookFile of requiredHooks) {
-      if (!existsSync(join(hooksDir, hookFile))) {
-        throw new Error(`Package hooks directory missing required file: ${hookFile}`);
+    const hookTemplatesDir = join(packageRoot, 'hook-templates');
+
+    // Verify it contains expected hook template files
+    const requiredHookTemplates = [
+      'session-start.js',
+      'stop.js',
+      'user-prompt-submit.js',
+      'post-tool-use.js'
+    ];
+    for (const hookTemplateFile of requiredHookTemplates) {
+      if (!existsSync(join(hookTemplatesDir, hookTemplateFile))) {
+        throw new Error(`Package hook-templates directory missing required template file: ${hookTemplateFile}`);
      }
    }
-    
-    return hooksDir;
+
+    return hookTemplatesDir;
  }

  /**
@@ -325,9 +335,19 @@ export class PathDiscovery {

  /**
   * Get current project directory name
+   * Uses git repository root's basename if in a git repo, otherwise falls back to cwd basename
   */
  static getCurrentProjectName(): string {
-    return require('path').basename(process.cwd());
+    try {
+      const gitRoot = execSync('git rev-parse --show-toplevel', {
+        cwd: process.cwd(),
+        encoding: 'utf8',
+        stdio: ['pipe', 'pipe', 'ignore']
+      }).trim();
+      return require('path').basename(gitRoot);
+    } catch {
+      return require('path').basename(process.cwd());
+    }
  }

  /**
@@ -347,68 +367,7 @@ export class PathDiscovery {
   * Check if a path exists and is accessible
   */
  static isPathAccessible(path: string): boolean {
-    try {
-      return existsSync(path) && statSync(path).isDirectory();
-    } catch {
-      return false;
-    }
+    return existsSync(path) && statSync(path).isDirectory();
  }

-  // =============================================================================
-  // STATIC CONVENIENCE METHODS
-  // =============================================================================
-
-  /**
-   * Quick access to singleton instance methods
-   */
-  static getDataDirectory(): string {
-    return PathDiscovery.getInstance().getDataDirectory();
-  }
-
-  static getArchivesDirectory(): string {
-    return PathDiscovery.getInstance().getArchivesDirectory();
-  }
-
-  static getHooksDirectory(): string {
-    return PathDiscovery.getInstance().getHooksDirectory();
-  }
-
-  static getLogsDirectory(): string {
-    return PathDiscovery.getInstance().getLogsDirectory();
-  }
-
-  static getClaudeSettingsPath(): string {
-    return PathDiscovery.getInstance().getClaudeSettingsPath();
-  }
-
-  static getClaudeMdPath(): string {
-    return PathDiscovery.getInstance().getClaudeMdPath();
-  }
-
-  static findPackageHooksDirectory(): string {
-    return PathDiscovery.getInstance().findPackageHooksDirectory();
-  }
-
-  static findPackageCommandsDirectory(): string {
-    return PathDiscovery.getInstance().findPackageCommandsDirectory();
-  }
-}
-
-// Export singleton instance for immediate use
-export const pathDiscovery = PathDiscovery.getInstance();
-
-// Export static methods for convenience
-export const {
-  getDataDirectory,
-  getArchivesDirectory, 
-  getHooksDirectory,
-  getLogsDirectory,
-  getClaudeSettingsPath,
-  getClaudeMdPath,
-  findPackageHooksDirectory,
-  findPackageCommandsDirectory,
-  extractProjectName,
-  getCurrentProjectName,
-  createBackupFilename,
-  isPathAccessible
-} = PathDiscovery;
+}
@@ -17,12 +17,12 @@ export class MemoryStore {
   */
  create(input: MemoryInput): MemoryRow {
    const { isoString, epoch } = normalizeTimestamp(input.created_at);
-    
+
    const stmt = this.db.prepare(`
      INSERT INTO memories (
-        session_id, text, document_id, keywords, created_at, created_at_epoch, 
-        project, archive_basename, origin
-      ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
+        session_id, text, document_id, keywords, created_at, created_at_epoch,
+        project, archive_basename, origin, title, subtitle, facts, concepts, files_touched
+      ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
    `);

    const info = stmt.run(
@@ -34,7 +34,12 @@ export class MemoryStore {
      epoch,
      input.project,
      input.archive_basename || null,
-      input.origin || 'transcript'
+      input.origin || 'transcript',
+      input.title || null,
+      input.subtitle || null,
+      input.facts || null,
+      input.concepts || null,
+      input.files_touched || null
    );

    return this.getById(info.lastInsertRowid as number)!;
@@ -158,15 +163,44 @@ export class MemoryStore {
   * Get memories by origin type
   */
  getByOrigin(origin: string, limit?: number): MemoryRow[] {
-    const query = limit 
+    const query = limit
      ? 'SELECT * FROM memories WHERE origin = ? ORDER BY created_at_epoch DESC LIMIT ?'
      : 'SELECT * FROM memories WHERE origin = ? ORDER BY created_at_epoch DESC';
-    
+
    const stmt = this.db.prepare(query);
    const params = limit ? [origin, limit] : [origin];
    return stmt.all(...params) as MemoryRow[];
  }

+  /**
+   * Get recent memories for a project filtered by origin
+   */
+  getRecentForProjectByOrigin(project: string, origin: string, limit = 10): MemoryRow[] {
+    const stmt = this.db.prepare(`
+      SELECT * FROM memories
+      WHERE project = ? AND origin = ?
+      ORDER BY created_at_epoch DESC
+      LIMIT ?
+    `);
+    return stmt.all(project, origin, limit) as MemoryRow[];
+  }
+
+  /**
+   * Get last N memories for a project, sorted oldest to newest
+   */
+  getLastNForProject(project: string, limit = 10): MemoryRow[] {
+    const stmt = this.db.prepare(`
+      SELECT * FROM (
+        SELECT * FROM memories
+        WHERE project = ?
+        ORDER BY created_at_epoch DESC
+        LIMIT ?
+      )
+      ORDER BY created_at_epoch ASC
+    `);
+    return stmt.all(project, limit) as MemoryRow[];
+  }
+
  /**
   * Count total memories
   */
@@ -195,11 +229,12 @@ export class MemoryStore {
    }

    const { isoString, epoch } = normalizeTimestamp(input.created_at || existing.created_at);
-    
+
    const stmt = this.db.prepare(`
      UPDATE memories SET
        text = ?, document_id = ?, keywords = ?, created_at = ?, created_at_epoch = ?,
-        project = ?, archive_basename = ?, origin = ?
+        project = ?, archive_basename = ?, origin = ?, title = ?, subtitle = ?, facts = ?,
+        concepts = ?, files_touched = ?
      WHERE id = ?
    `);

@@ -212,6 +247,11 @@ export class MemoryStore {
      input.project || existing.project,
      input.archive_basename !== undefined ? input.archive_basename : existing.archive_basename,
      input.origin || existing.origin,
+      input.title !== undefined ? input.title : existing.title,
+      input.subtitle !== undefined ? input.subtitle : existing.subtitle,
+      input.facts !== undefined ? input.facts : existing.facts,
+      input.concepts !== undefined ? input.concepts : existing.concepts,
+      input.files_touched !== undefined ? input.files_touched : existing.files_touched,
      id
    );

@@ -68,14 +68,26 @@ export class OverviewStore {
   */
  getRecentForProject(project: string, limit = 5): OverviewRow[] {
    const stmt = this.db.prepare(`
-      SELECT * FROM overviews 
+      SELECT * FROM overviews
      WHERE project = ?
-      ORDER BY created_at_epoch DESC 
+      ORDER BY created_at_epoch DESC
      LIMIT ?
    `);
    return stmt.all(project, limit) as OverviewRow[];
  }

+  /**
+   * Get all overviews for a project (oldest to newest)
+   */
+  getAllForProject(project: string): OverviewRow[] {
+    const stmt = this.db.prepare(`
+      SELECT * FROM overviews
+      WHERE project = ?
+      ORDER BY created_at_epoch ASC
+    `);
+    return stmt.all(project) as OverviewRow[];
+  }
+
  /**
   * Get recent overviews across all projects
   */
@@ -193,4 +205,37 @@ export class OverviewStore {
    const rows = stmt.all() as { project: string }[];
    return rows.map(row => row.project);
  }
+
+  /**
+   * Get most recent overview for a specific project
+   */
+  getByProject(project: string): OverviewRow | null {
+    const stmt = this.db.prepare(`
+      SELECT * FROM overviews
+      WHERE project = ?
+      ORDER BY created_at_epoch DESC
+      LIMIT 1
+    `);
+    return stmt.get(project) as OverviewRow || null;
+  }
+
+  /**
+   * Create or update overview for a project (keeps only most recent)
+   */
+  upsertByProject(input: OverviewInput): OverviewRow {
+    const existing = this.getByProject(input.project);
+    if (existing) {
+      return this.update(existing.id, input);
+    }
+    return this.create(input);
+  }
+
+  /**
+   * Delete overview by project name
+   */
+  deleteByProject(project: string): boolean {
+    const stmt = this.db.prepare('DELETE FROM overviews WHERE project = ?');
+    const info = stmt.run(project);
+    return info.changes > 0;
+  }
 }
@@ -0,0 +1,266 @@
+import { Database } from 'better-sqlite3';
+import { getDatabase } from './Database.js';
+import { normalizeTimestamp } from './types.js';
+
+/**
+ * Represents a streaming session row in the database
+ */
+export interface StreamingSessionRow {
+  id: number;
+  claude_session_id: string;
+  sdk_session_id?: string;
+  project: string;
+  title?: string;
+  subtitle?: string;
+  user_prompt?: string;
+  started_at: string;
+  started_at_epoch: number;
+  updated_at?: string;
+  updated_at_epoch?: number;
+  completed_at?: string;
+  completed_at_epoch?: number;
+  status: 'active' | 'completed' | 'failed';
+}
+
+/**
+ * Input type for creating a new streaming session
+ */
+export interface StreamingSessionInput {
+  claude_session_id: string;
+  project: string;
+  user_prompt?: string;
+  started_at?: string | Date | number;
+}
+
+/**
+ * Input type for updating a streaming session
+ */
+export interface StreamingSessionUpdate {
+  sdk_session_id?: string;
+  title?: string;
+  subtitle?: string;
+  status?: 'active' | 'completed' | 'failed';
+  completed_at?: string | Date | number;
+}
+
+/**
+ * Data Access Object for streaming session records
+ * Handles real-time session tracking during SDK compression
+ */
+export class StreamingSessionStore {
+  private db: Database.Database;
+
+  constructor(db?: Database.Database) {
+    this.db = db || getDatabase();
+  }
+
+  /**
+   * Create a new streaming session record
+   * This should be called immediately when the hook receives a user prompt
+   */
+  create(input: StreamingSessionInput): StreamingSessionRow {
+    const { isoString, epoch } = normalizeTimestamp(input.started_at);
+
+    const stmt = this.db.prepare(`
+      INSERT INTO streaming_sessions (
+        claude_session_id, project, user_prompt, started_at, started_at_epoch, status
+      ) VALUES (?, ?, ?, ?, ?, 'active')
+    `);
+
+    const info = stmt.run(
+      input.claude_session_id,
+      input.project,
+      input.user_prompt || null,
+      isoString,
+      epoch
+    );
+
+    return this.getById(info.lastInsertRowid as number)!;
+  }
+
+  /**
+   * Update a streaming session by internal ID
+   * Uses atomic transaction to prevent race conditions
+   */
+  update(id: number, updates: StreamingSessionUpdate): StreamingSessionRow {
+    const { isoString: updatedAt, epoch: updatedEpoch } = normalizeTimestamp(new Date());
+
+    const existing = this.getById(id);
+    if (!existing) {
+      throw new Error(`Streaming session with id ${id} not found`);
+    }
+
+    const parts: string[] = [];
+    const values: any[] = [];
+
+    if (updates.sdk_session_id !== undefined) {
+      parts.push('sdk_session_id = ?');
+      values.push(updates.sdk_session_id);
+    }
+    if (updates.title !== undefined) {
+      parts.push('title = ?');
+      values.push(updates.title);
+    }
+    if (updates.subtitle !== undefined) {
+      parts.push('subtitle = ?');
+      values.push(updates.subtitle);
+    }
+    if (updates.status !== undefined) {
+      parts.push('status = ?');
+      values.push(updates.status);
+    }
+    if (updates.completed_at !== undefined) {
+      const { isoString, epoch } = normalizeTimestamp(updates.completed_at);
+      parts.push('completed_at = ?', 'completed_at_epoch = ?');
+      values.push(isoString, epoch);
+    }
+
+    // Always update the updated_at timestamp
+    parts.push('updated_at = ?', 'updated_at_epoch = ?');
+    values.push(updatedAt, updatedEpoch);
+
+    values.push(id);
+
+    const stmt = this.db.prepare(`
+      UPDATE streaming_sessions
+      SET ${parts.join(', ')}
+      WHERE id = ?
+    `);
+
+    stmt.run(...values);
+
+    return this.getById(id)!;
+  }
+
+  /**
+   * Update a streaming session by Claude session ID
+   * Convenience method for hooks that only have the Claude session ID
+   */
+  updateByClaudeSessionId(claudeSessionId: string, updates: StreamingSessionUpdate): StreamingSessionRow | null {
+    const session = this.getByClaudeSessionId(claudeSessionId);
+    if (!session) {
+      return null;
+    }
+    return this.update(session.id, updates);
+  }
+
+  /**
+   * Get streaming session by internal ID
+   */
+  getById(id: number): StreamingSessionRow | null {
+    const stmt = this.db.prepare('SELECT * FROM streaming_sessions WHERE id = ?');
+    return stmt.get(id) as StreamingSessionRow || null;
+  }
+
+  /**
+   * Get streaming session by Claude session ID
+   */
+  getByClaudeSessionId(claudeSessionId: string): StreamingSessionRow | null {
+    const stmt = this.db.prepare('SELECT * FROM streaming_sessions WHERE claude_session_id = ?');
+    return stmt.get(claudeSessionId) as StreamingSessionRow || null;
+  }
+
+  /**
+   * Get streaming session by SDK session ID
+   */
+  getBySdkSessionId(sdkSessionId: string): StreamingSessionRow | null {
+    const stmt = this.db.prepare('SELECT * FROM streaming_sessions WHERE sdk_session_id = ?');
+    return stmt.get(sdkSessionId) as StreamingSessionRow || null;
+  }
+
+  /**
+   * Check if a streaming session exists by Claude session ID
+   */
+  has(claudeSessionId: string): boolean {
+    const stmt = this.db.prepare('SELECT 1 FROM streaming_sessions WHERE claude_session_id = ? LIMIT 1');
+    return Boolean(stmt.get(claudeSessionId));
+  }
+
+  /**
+   * Get active streaming sessions for a project
+   */
+  getActiveForProject(project: string): StreamingSessionRow[] {
+    const stmt = this.db.prepare(`
+      SELECT * FROM streaming_sessions
+      WHERE project = ? AND status = 'active'
+      ORDER BY started_at_epoch DESC
+    `);
+    return stmt.all(project) as StreamingSessionRow[];
+  }
+
+  /**
+   * Get all active streaming sessions
+   */
+  getAllActive(): StreamingSessionRow[] {
+    const stmt = this.db.prepare(`
+      SELECT * FROM streaming_sessions
+      WHERE status = 'active'
+      ORDER BY started_at_epoch DESC
+    `);
+    return stmt.all() as StreamingSessionRow[];
+  }
+
+  /**
+   * Get recent streaming sessions (completed or failed)
+   */
+  getRecent(limit = 10): StreamingSessionRow[] {
+    const stmt = this.db.prepare(`
+      SELECT * FROM streaming_sessions
+      ORDER BY started_at_epoch DESC
+      LIMIT ?
+    `);
+    return stmt.all(limit) as StreamingSessionRow[];
+  }
+
+  /**
+   * Mark a session as completed
+   */
+  markCompleted(id: number): StreamingSessionRow {
+    return this.update(id, {
+      status: 'completed',
+      completed_at: new Date()
+    });
+  }
+
+  /**
+   * Mark a session as failed
+   */
+  markFailed(id: number): StreamingSessionRow {
+    return this.update(id, {
+      status: 'failed',
+      completed_at: new Date()
+    });
+  }
+
+  /**
+   * Delete a streaming session by ID
+   */
+  deleteById(id: number): boolean {
+    const stmt = this.db.prepare('DELETE FROM streaming_sessions WHERE id = ?');
+    const info = stmt.run(id);
+    return info.changes > 0;
+  }
+
+  /**
+   * Delete a streaming session by Claude session ID
+   */
+  deleteByClaudeSessionId(claudeSessionId: string): boolean {
+    const stmt = this.db.prepare('DELETE FROM streaming_sessions WHERE claude_session_id = ?');
+    const info = stmt.run(claudeSessionId);
+    return info.changes > 0;
+  }
+
+  /**
+   * Clean up old completed/failed sessions (older than N days)
+   */
+  cleanupOldSessions(daysOld = 30): number {
+    const cutoffEpoch = Date.now() - (daysOld * 24 * 60 * 60 * 1000);
+    const stmt = this.db.prepare(`
+      DELETE FROM streaming_sessions
+      WHERE status IN ('completed', 'failed')
+        AND completed_at_epoch < ?
+    `);
+    const info = stmt.run(cutoffEpoch);
+    return info.changes;
+  }
+}
@@ -0,0 +1,107 @@
+import { Database } from 'better-sqlite3';
+import { getDatabase } from './Database.js';
+import {
+  TranscriptEventInput,
+  TranscriptEventRow,
+  normalizeTimestamp
+} from './types.js';
+
+/**
+ * Data access for transcript_events table
+ */
+export class TranscriptEventStore {
+  private db: Database.Database;
+
+  constructor(db?: Database.Database) {
+    this.db = db || getDatabase();
+  }
+
+  /**
+   * Insert or update a transcript event
+   */
+  upsert(event: TranscriptEventInput): TranscriptEventRow {
+    const { isoString, epoch } = normalizeTimestamp(event.captured_at);
+
+    const stmt = this.db.prepare(`
+      INSERT INTO transcript_events (
+        session_id,
+        project,
+        event_index,
+        event_type,
+        raw_json,
+        captured_at,
+        captured_at_epoch
+      ) VALUES (?, ?, ?, ?, ?, ?, ?)
+      ON CONFLICT(session_id, event_index) DO UPDATE SET
+        project = excluded.project,
+        event_type = excluded.event_type,
+        raw_json = excluded.raw_json,
+        captured_at = excluded.captured_at,
+        captured_at_epoch = excluded.captured_at_epoch
+    `);
+
+    stmt.run(
+      event.session_id,
+      event.project || null,
+      event.event_index,
+      event.event_type || null,
+      event.raw_json,
+      isoString,
+      epoch
+    );
+
+    return this.getBySessionAndIndex(event.session_id, event.event_index)!;
+  }
+
+  /**
+   * Bulk upsert events in a single transaction
+   */
+  upsertMany(events: TranscriptEventInput[]): TranscriptEventRow[] {
+    const transaction = this.db.transaction((rows: TranscriptEventInput[]) => {
+      const results: TranscriptEventRow[] = [];
+      for (const row of rows) {
+        results.push(this.upsert(row));
+      }
+      return results;
+    });
+
+    return transaction(events);
+  }
+
+  /**
+   * Get event by session and index
+   */
+  getBySessionAndIndex(sessionId: string, eventIndex: number): TranscriptEventRow | null {
+    const stmt = this.db.prepare(`
+      SELECT * FROM transcript_events
+      WHERE session_id = ? AND event_index = ?
+    `);
+    return stmt.get(sessionId, eventIndex) as TranscriptEventRow | null;
+  }
+
+  /**
+   * Get highest event_index stored for a session
+   */
+  getMaxEventIndex(sessionId: string): number {
+    const stmt = this.db.prepare(`
+      SELECT MAX(event_index) as max_event_index
+      FROM transcript_events
+      WHERE session_id = ?
+    `);
+    const row = stmt.get(sessionId) as { max_event_index: number | null } | undefined;
+    return row?.max_event_index ?? -1;
+  }
+
+  /**
+   * List recent events for a session
+   */
+  listBySession(sessionId: string, limit = 200, offset = 0): TranscriptEventRow[] {
+    const stmt = this.db.prepare(`
+      SELECT * FROM transcript_events
+      WHERE session_id = ?
+      ORDER BY event_index ASC
+      LIMIT ? OFFSET ?
+    `);
+    return stmt.all(sessionId, limit, offset) as TranscriptEventRow[];
+  }
+}
@@ -1,6 +1,3 @@
-// Import migrations to register them
-import './migrations/index.js';
-
 // Export main components
 export { DatabaseManager, getDatabase, initializeDatabase } from './Database.js';

@@ -9,24 +6,41 @@ export { SessionStore } from './SessionStore.js';
 export { MemoryStore } from './MemoryStore.js';
 export { OverviewStore } from './OverviewStore.js';
 export { DiagnosticsStore } from './DiagnosticsStore.js';
+export { TranscriptEventStore } from './TranscriptEventStore.js';
+export { StreamingSessionStore } from './StreamingSessionStore.js';

 // Export types
 export * from './types.js';

+// Export migrations
+export { migrations } from './migrations.js';
+
 // Convenience function to get all stores
 export async function createStores() {
  const { DatabaseManager } = await import('./Database.js');
-  const db = await DatabaseManager.getInstance().initialize();
-  
+  const { migrations } = await import('./migrations.js');
+
+  // Register migrations before initialization
+  const manager = DatabaseManager.getInstance();
+  for (const migration of migrations) {
+    manager.registerMigration(migration);
+  }
+
+  const db = await manager.initialize();
+
  const { SessionStore } = await import('./SessionStore.js');
  const { MemoryStore } = await import('./MemoryStore.js');
  const { OverviewStore } = await import('./OverviewStore.js');
  const { DiagnosticsStore } = await import('./DiagnosticsStore.js');
-  
+  const { TranscriptEventStore } = await import('./TranscriptEventStore.js');
+  const { StreamingSessionStore } = await import('./StreamingSessionStore.js');
+
  return {
    sessions: new SessionStore(db),
    memories: new MemoryStore(db),
    overviews: new OverviewStore(db),
-    diagnostics: new DiagnosticsStore(db)
+    diagnostics: new DiagnosticsStore(db),
+    transcriptEvents: new TranscriptEventStore(db),
+    streamingSessions: new StreamingSessionStore(db)
  };
 }
@@ -0,0 +1,212 @@
+import { Database } from 'better-sqlite3';
+import { Migration } from './Database.js';
+
+/**
+ * Initial schema migration - creates all core tables
+ */
+export const migration001: Migration = {
+  version: 1,
+  up: (db: Database.Database) => {
+    // Sessions table - core session tracking
+    db.exec(`
+      CREATE TABLE IF NOT EXISTS sessions (
+        id INTEGER PRIMARY KEY AUTOINCREMENT,
+        session_id TEXT UNIQUE NOT NULL,
+        project TEXT NOT NULL,
+        created_at TEXT NOT NULL,
+        created_at_epoch INTEGER NOT NULL,
+        source TEXT NOT NULL DEFAULT 'compress',
+        archive_path TEXT,
+        archive_bytes INTEGER,
+        archive_checksum TEXT,
+        archived_at TEXT,
+        metadata_json TEXT
+      );
+
+      CREATE INDEX IF NOT EXISTS idx_sessions_project ON sessions(project);
+      CREATE INDEX IF NOT EXISTS idx_sessions_created_at ON sessions(created_at_epoch DESC);
+      CREATE INDEX IF NOT EXISTS idx_sessions_project_created ON sessions(project, created_at_epoch DESC);
+    `);
+
+    // Memories table - compressed memory chunks
+    db.exec(`
+      CREATE TABLE IF NOT EXISTS memories (
+        id INTEGER PRIMARY KEY AUTOINCREMENT,
+        session_id TEXT NOT NULL,
+        text TEXT NOT NULL,
+        document_id TEXT UNIQUE,
+        keywords TEXT,
+        created_at TEXT NOT NULL,
+        created_at_epoch INTEGER NOT NULL,
+        project TEXT NOT NULL,
+        archive_basename TEXT,
+        origin TEXT NOT NULL DEFAULT 'transcript',
+        FOREIGN KEY (session_id) REFERENCES sessions(session_id) ON DELETE CASCADE
+      );
+
+      CREATE INDEX IF NOT EXISTS idx_memories_session ON memories(session_id);
+      CREATE INDEX IF NOT EXISTS idx_memories_project ON memories(project);
+      CREATE INDEX IF NOT EXISTS idx_memories_created_at ON memories(created_at_epoch DESC);
+      CREATE INDEX IF NOT EXISTS idx_memories_project_created ON memories(project, created_at_epoch DESC);
+      CREATE INDEX IF NOT EXISTS idx_memories_document_id ON memories(document_id);
+      CREATE INDEX IF NOT EXISTS idx_memories_origin ON memories(origin);
+    `);
+
+    // Overviews table - session summaries (one per project)
+    db.exec(`
+      CREATE TABLE IF NOT EXISTS overviews (
+        id INTEGER PRIMARY KEY AUTOINCREMENT,
+        session_id TEXT NOT NULL,
+        content TEXT NOT NULL,
+        created_at TEXT NOT NULL,
+        created_at_epoch INTEGER NOT NULL,
+        project TEXT NOT NULL,
+        origin TEXT NOT NULL DEFAULT 'claude',
+        FOREIGN KEY (session_id) REFERENCES sessions(session_id) ON DELETE CASCADE
+      );
+
+      CREATE INDEX IF NOT EXISTS idx_overviews_session ON overviews(session_id);
+      CREATE INDEX IF NOT EXISTS idx_overviews_project ON overviews(project);
+      CREATE INDEX IF NOT EXISTS idx_overviews_created_at ON overviews(created_at_epoch DESC);
+      CREATE INDEX IF NOT EXISTS idx_overviews_project_created ON overviews(project, created_at_epoch DESC);
+      CREATE UNIQUE INDEX IF NOT EXISTS idx_overviews_project_latest ON overviews(project, created_at_epoch DESC);
+    `);
+
+    // Diagnostics table - system health and debug info
+    db.exec(`
+      CREATE TABLE IF NOT EXISTS diagnostics (
+        id INTEGER PRIMARY KEY AUTOINCREMENT,
+        session_id TEXT,
+        message TEXT NOT NULL,
+        severity TEXT NOT NULL DEFAULT 'info',
+        created_at TEXT NOT NULL,
+        created_at_epoch INTEGER NOT NULL,
+        project TEXT NOT NULL,
+        origin TEXT NOT NULL DEFAULT 'system',
+        FOREIGN KEY (session_id) REFERENCES sessions(session_id) ON DELETE SET NULL
+      );
+
+      CREATE INDEX IF NOT EXISTS idx_diagnostics_session ON diagnostics(session_id);
+      CREATE INDEX IF NOT EXISTS idx_diagnostics_project ON diagnostics(project);
+      CREATE INDEX IF NOT EXISTS idx_diagnostics_severity ON diagnostics(severity);
+      CREATE INDEX IF NOT EXISTS idx_diagnostics_created ON diagnostics(created_at_epoch DESC);
+    `);
+
+    // Transcript events table - raw conversation events
+    db.exec(`
+      CREATE TABLE IF NOT EXISTS transcript_events (
+        id INTEGER PRIMARY KEY AUTOINCREMENT,
+        session_id TEXT NOT NULL,
+        project TEXT,
+        event_index INTEGER NOT NULL,
+        event_type TEXT,
+        raw_json TEXT NOT NULL,
+        captured_at TEXT NOT NULL,
+        captured_at_epoch INTEGER NOT NULL,
+        FOREIGN KEY (session_id) REFERENCES sessions(session_id) ON DELETE CASCADE,
+        UNIQUE(session_id, event_index)
+      );
+
+      CREATE INDEX IF NOT EXISTS idx_transcript_events_session ON transcript_events(session_id, event_index);
+      CREATE INDEX IF NOT EXISTS idx_transcript_events_project ON transcript_events(project);
+      CREATE INDEX IF NOT EXISTS idx_transcript_events_type ON transcript_events(event_type);
+      CREATE INDEX IF NOT EXISTS idx_transcript_events_captured ON transcript_events(captured_at_epoch DESC);
+    `);
+
+    console.log('✅ Created all database tables successfully');
+  },
+
+  down: (db: Database.Database) => {
+    db.exec(`
+      DROP TABLE IF EXISTS transcript_events;
+      DROP TABLE IF EXISTS diagnostics;
+      DROP TABLE IF EXISTS overviews;
+      DROP TABLE IF EXISTS memories;
+      DROP TABLE IF EXISTS sessions;
+    `);
+  }
+};
+
+/**
+ * Migration 002 - Add hierarchical memory fields (v2 format)
+ */
+export const migration002: Migration = {
+  version: 2,
+  up: (db: Database.Database) => {
+    // Add new columns for hierarchical memory structure
+    db.exec(`
+      ALTER TABLE memories ADD COLUMN title TEXT;
+      ALTER TABLE memories ADD COLUMN subtitle TEXT;
+      ALTER TABLE memories ADD COLUMN facts TEXT;
+      ALTER TABLE memories ADD COLUMN concepts TEXT;
+      ALTER TABLE memories ADD COLUMN files_touched TEXT;
+    `);
+
+    // Create indexes for the new fields to improve search performance
+    db.exec(`
+      CREATE INDEX IF NOT EXISTS idx_memories_title ON memories(title);
+      CREATE INDEX IF NOT EXISTS idx_memories_concepts ON memories(concepts);
+    `);
+
+    console.log('✅ Added hierarchical memory fields to memories table');
+  },
+
+  down: (db: Database.Database) => {
+    // Note: SQLite doesn't support DROP COLUMN in all versions
+    // In production, we'd need to recreate the table without these columns
+    // For now, we'll just log a warning
+    console.log('⚠️  Warning: SQLite ALTER TABLE DROP COLUMN not fully supported');
+    console.log('⚠️  To rollback, manually recreate the memories table');
+  }
+};
+
+/**
+ * Migration 003 - Add streaming_sessions table for real-time session tracking
+ */
+export const migration003: Migration = {
+  version: 3,
+  up: (db: Database.Database) => {
+    // Streaming sessions table - tracks active SDK compression sessions
+    db.exec(`
+      CREATE TABLE IF NOT EXISTS streaming_sessions (
+        id INTEGER PRIMARY KEY AUTOINCREMENT,
+        claude_session_id TEXT UNIQUE NOT NULL,
+        sdk_session_id TEXT,
+        project TEXT NOT NULL,
+        title TEXT,
+        subtitle TEXT,
+        user_prompt TEXT,
+        started_at TEXT NOT NULL,
+        started_at_epoch INTEGER NOT NULL,
+        updated_at TEXT,
+        updated_at_epoch INTEGER,
+        completed_at TEXT,
+        completed_at_epoch INTEGER,
+        status TEXT NOT NULL DEFAULT 'active'
+      );
+
+      CREATE INDEX IF NOT EXISTS idx_streaming_sessions_claude_id ON streaming_sessions(claude_session_id);
+      CREATE INDEX IF NOT EXISTS idx_streaming_sessions_sdk_id ON streaming_sessions(sdk_session_id);
+      CREATE INDEX IF NOT EXISTS idx_streaming_sessions_project ON streaming_sessions(project);
+      CREATE INDEX IF NOT EXISTS idx_streaming_sessions_status ON streaming_sessions(status);
+      CREATE INDEX IF NOT EXISTS idx_streaming_sessions_started ON streaming_sessions(started_at_epoch DESC);
+    `);
+
+    console.log('✅ Created streaming_sessions table for real-time session tracking');
+  },
+
+  down: (db: Database.Database) => {
+    db.exec(`
+      DROP TABLE IF EXISTS streaming_sessions;
+    `);
+  }
+};
+
+/**
+ * All migrations in order
+ */
+export const migrations: Migration[] = [
+  migration001,
+  migration002,
+  migration003
+];
@@ -1,133 +0,0 @@
-import { Migration } from '../Database.js';
-
-/**
- * Initial migration: Create all core tables for claude-mem SQLite index
- */
-export const migration001: Migration = {
-  version: 1,
-  
-  up: (db) => {
-    // Create sessions table
-    db.exec(`
-      CREATE TABLE sessions (
-        id INTEGER PRIMARY KEY AUTOINCREMENT,
-        session_id TEXT UNIQUE NOT NULL,
-        project TEXT NOT NULL,
-        created_at TEXT NOT NULL,
-        created_at_epoch INTEGER NOT NULL,
-        source TEXT DEFAULT 'compress',
-        archive_path TEXT,
-        archive_bytes INTEGER,
-        archive_checksum TEXT,
-        archived_at TEXT,
-        metadata_json TEXT
-      )
-    `);
-
-    // Create indexes for sessions
-    db.exec(`
-      CREATE INDEX sessions_project_created_at ON sessions (project, created_at_epoch DESC)
-    `);
-    db.exec(`
-      CREATE INDEX sessions_source_created ON sessions (source, created_at_epoch DESC)
-    `);
-
-    // Create overviews table
-    db.exec(`
-      CREATE TABLE overviews (
-        id INTEGER PRIMARY KEY AUTOINCREMENT,
-        session_id TEXT NOT NULL REFERENCES sessions(session_id) ON DELETE CASCADE,
-        content TEXT NOT NULL,
-        created_at TEXT NOT NULL,
-        created_at_epoch INTEGER NOT NULL,
-        project TEXT NOT NULL,
-        origin TEXT DEFAULT 'claude'
-      )
-    `);
-
-    // Create index for overviews
-    db.exec(`
-      CREATE INDEX overviews_project_created_at ON overviews (project, created_at_epoch DESC)
-    `);
-
-    // Create memories table
-    db.exec(`
-      CREATE TABLE memories (
-        id INTEGER PRIMARY KEY AUTOINCREMENT,
-        session_id TEXT NOT NULL REFERENCES sessions(session_id) ON DELETE CASCADE,
-        text TEXT NOT NULL,
-        document_id TEXT,
-        keywords TEXT,
-        created_at TEXT NOT NULL,
-        created_at_epoch INTEGER NOT NULL,
-        project TEXT NOT NULL,
-        archive_basename TEXT,
-        origin TEXT DEFAULT 'transcript'
-      )
-    `);
-
-    // Create indexes for memories
-    db.exec(`
-      CREATE INDEX memories_project_created_at ON memories (project, created_at_epoch DESC)
-    `);
-    db.exec(`
-      CREATE UNIQUE INDEX memories_document_id_unique ON memories (document_id) WHERE document_id IS NOT NULL
-    `);
-
-    // Create diagnostics table
-    db.exec(`
-      CREATE TABLE diagnostics (
-        id INTEGER PRIMARY KEY AUTOINCREMENT,
-        session_id TEXT REFERENCES sessions(session_id) ON DELETE SET NULL,
-        message TEXT NOT NULL,
-        severity TEXT DEFAULT 'warn',
-        created_at TEXT NOT NULL,
-        created_at_epoch INTEGER NOT NULL,
-        project TEXT NOT NULL,
-        origin TEXT DEFAULT 'compressor'
-      )
-    `);
-
-    // Create index for diagnostics
-    db.exec(`
-      CREATE INDEX diagnostics_project_created_at ON diagnostics (project, created_at_epoch DESC)
-    `);
-
-    // Create archives table (for future archival workflows)
-    db.exec(`
-      CREATE TABLE archives (
-        id INTEGER PRIMARY KEY AUTOINCREMENT,
-        session_id TEXT UNIQUE NOT NULL REFERENCES sessions(session_id) ON DELETE CASCADE,
-        path TEXT NOT NULL,
-        bytes INTEGER,
-        checksum TEXT,
-        stored_at TEXT NOT NULL,
-        storage_status TEXT DEFAULT 'active'
-      )
-    `);
-
-    // Create titles table (ready for conversation-titles.jsonl migration)
-    db.exec(`
-      CREATE TABLE titles (
-        id INTEGER PRIMARY KEY AUTOINCREMENT,
-        session_id TEXT UNIQUE NOT NULL REFERENCES sessions(session_id) ON DELETE CASCADE,
-        title TEXT NOT NULL,
-        created_at TEXT NOT NULL,
-        project TEXT NOT NULL
-      )
-    `);
-
-    console.log('✅ Created initial database schema with all tables and indexes');
-  },
-
-  down: (db) => {
-    // Drop tables in reverse order to respect foreign key constraints
-    const tables = ['titles', 'archives', 'diagnostics', 'memories', 'overviews', 'sessions'];
-    
-    for (const table of tables) {
-      db.exec(`DROP TABLE IF EXISTS ${table}`);
-    }
-    
-    console.log('🗑️ Dropped all tables from initial migration');
-  }
-};
@@ -1,17 +0,0 @@
-import { DatabaseManager } from '../Database.js';
-import { migration001 } from './001_initial.js';
-
-/**
- * Register all migrations with the database manager
- */
-export function registerMigrations(): void {
-  const manager = DatabaseManager.getInstance();
-  
-  // Register migrations in order
-  manager.registerMigration(migration001);
-  
-  console.log('📋 Registered all database migrations');
-}
-
-// Auto-register migrations when this module is imported
-registerMigrations();
@@ -37,6 +37,12 @@ export interface MemoryRow {
  project: string;
  archive_basename?: string;
  origin: string;
+  // Hierarchical memory fields (v2)
+  title?: string;
+  subtitle?: string;
+  facts?: string; // JSON array of fact strings
+  concepts?: string; // JSON array of concept strings
+  files_touched?: string; // JSON array of file paths
 }

 export interface DiagnosticRow {
@@ -50,6 +56,17 @@ export interface DiagnosticRow {
  origin: string;
 }

+export interface TranscriptEventRow {
+  id: number;
+  session_id: string;
+  project?: string;
+  event_index: number;
+  event_type?: string;
+  raw_json: string;
+  captured_at: string;
+  captured_at_epoch: number;
+}
+
 export interface ArchiveRow {
  id: number;
  session_id: string;
@@ -100,6 +117,12 @@ export interface MemoryInput {
  project: string;
  archive_basename?: string;
  origin?: string;
+  // Hierarchical memory fields (v2)
+  title?: string;
+  subtitle?: string;
+  facts?: string; // JSON array of fact strings
+  concepts?: string; // JSON array of concept strings
+  files_touched?: string; // JSON array of file paths
 }

 export interface DiagnosticInput {
@@ -111,6 +134,15 @@ export interface DiagnosticInput {
  origin?: string;
 }

+export interface TranscriptEventInput {
+  session_id: string;
+  project?: string;
+  event_index: number;
+  event_type?: string;
+  raw_json: string;
+  captured_at?: string | Date | number;
+}
+
 /**
 * Helper function to normalize timestamps from various formats
 */
@@ -149,4 +181,4 @@ export function normalizeTimestamp(timestamp: string | Date | number | undefined
    isoString: date.toISOString(),
    epoch: date.getTime()
  };
-}
+}
@@ -1,218 +0,0 @@
-import fs from 'fs';
-import path from 'path';
-import { log } from '../shared/logger.js';
-import { PathDiscovery } from './path-discovery.js';
-
-/**
- * Interface for Claude Code JSONL conversation entries
- */
-export interface ClaudeCodeMessage {
-  sessionId: string;
-  timestamp: string;
-  gitBranch?: string;
-  cwd: string;
-  type: 'user' | 'assistant' | 'system' | 'result';
-  message: {
-    role: string;
-    content: Array<{
-      type: string;
-      text?: string;
-      thinking?: string;
-    }> | string;
-  };
-  uuid: string;
-  version?: string;
-  isSidechain?: boolean;
-  userType?: string;
-  parentUuid?: string;
-  subtype?: string;
-  model?: string;
-  stop_reason?: string;
-  usage?: any;
-}
-
-/**
- * Interface matching TranscriptCompressor's expected format
- */
-export interface TranscriptMessage {
-  type: string;
-  message?: {
-    content?: string | Array<{
-      text?: string;
-      content?: string;
-    }>;
-    role?: string;
-    timestamp?: string;
-    created_at?: string;
-  };
-  content?: string | Array<{
-    text?: string;
-    content?: string;
-  }>;
-  role?: string;
-  uuid?: string;
-  session_id?: string;
-  timestamp?: string;
-  created_at?: string;
-  subtype?: string;
-}
-
-/**
- * Parsed conversation with metadata
- */
-export interface ParsedConversation {
-  sessionId: string;
-  filePath: string;
-  messageCount: number;
-  timestamp: string;
-  gitBranch?: string;
-  cwd: string;
-  messages: TranscriptMessage[];
-}
-
-/**
- * Service for parsing Claude Code JSONL conversation files
- */
-export class TranscriptParser {
-  
-  /**
-   * Parse a single JSONL conversation file
-   */
-  async parseConversation(filePath: string): Promise<ParsedConversation> {
-    const content = fs.readFileSync(filePath, 'utf-8');
-    const lines = content.trim().split('\n').filter(line => line.trim());
-    
-    const claudeMessages: ClaudeCodeMessage[] = [];
-    let parseErrors = 0;
-    
-    for (let i = 0; i < lines.length; i++) {
-      try {
-        const parsed = JSON.parse(lines[i]);
-        claudeMessages.push(parsed);
-      } catch (e) {
-        parseErrors++;
-        log.debug(`Parse error on line ${i + 1}: ${(e as Error).message}`);
-      }
-    }
-    
-    if (claudeMessages.length === 0) {
-      throw new Error(`No valid messages found in ${filePath}`);
-    }
-    
-    // Get metadata from first message
-    const firstMessage = claudeMessages[0];
-    const sessionId = firstMessage.sessionId;
-    const timestamp = firstMessage.timestamp;
-    const gitBranch = firstMessage.gitBranch;
-    const cwd = firstMessage.cwd;
-    
-    // Convert to TranscriptMessage format
-    const messages = claudeMessages.map(msg => this.convertMessage(msg));
-    
-    log.debug(`Parsed ${filePath}: ${messages.length} messages, ${parseErrors} errors`);
-    
-    return {
-      sessionId,
-      filePath,
-      messageCount: messages.length,
-      timestamp,
-      gitBranch,
-      cwd,
-      messages
-    };
-  }
-  
-  /**
-   * Convert ClaudeCodeMessage to TranscriptMessage format
-   */
-  private convertMessage(claudeMsg: ClaudeCodeMessage): TranscriptMessage {
-    const converted: TranscriptMessage = {
-      type: claudeMsg.type,
-      uuid: claudeMsg.uuid,
-      session_id: claudeMsg.sessionId,
-      timestamp: claudeMsg.timestamp,
-      subtype: claudeMsg.subtype
-    };
-    
-    // Handle message content
-    if (claudeMsg.message) {
-      converted.message = {
-        role: claudeMsg.message.role,
-        timestamp: claudeMsg.timestamp
-      };
-      
-      if (Array.isArray(claudeMsg.message.content)) {
-        // Convert content array to expected format
-        converted.message.content = claudeMsg.message.content.map(item => ({
-          text: item.text || item.thinking || '',
-          content: item.text || item.thinking || ''
-        }));
-      } else if (typeof claudeMsg.message.content === 'string') {
-        converted.message.content = claudeMsg.message.content;
-      }
-    }
-    
-    return converted;
-  }
-  
-  /**
-   * Scan Claude projects directory for conversation files
-   */
-  async scanConversationFiles(): Promise<string[]> {
-    const pathDiscovery = PathDiscovery.getInstance();
-    const claudeDir = path.join(pathDiscovery.getClaudeConfigDirectory(), 'projects');
-    
-    if (!fs.existsSync(claudeDir)) {
-      return [];
-    }
-    
-    const projectDirs = fs.readdirSync(claudeDir);
-    const conversationFiles: string[] = [];
-    
-    for (const projectDir of projectDirs) {
-      const projectPath = path.join(claudeDir, projectDir);
-      if (!fs.statSync(projectPath).isDirectory()) continue;
-      
-      const files = fs.readdirSync(projectPath);
-      for (const file of files) {
-        if (file.endsWith('.jsonl')) {
-          conversationFiles.push(path.join(projectPath, file));
-        }
-      }
-    }
-    
-    return conversationFiles;
-  }
-  
-  /**
-   * Get conversation metadata without fully parsing
-   */
-  async getConversationMetadata(filePath: string): Promise<{
-    sessionId: string;
-    timestamp: string;
-    messageCount: number;
-    gitBranch?: string;
-    cwd: string;
-    fileSize: number;
-  }> {
-    const stats = fs.statSync(filePath);
-    const content = fs.readFileSync(filePath, 'utf-8');
-    const lines = content.trim().split('\n').filter(line => line.trim());
-    
-    let firstMessage;
-    try {
-      firstMessage = JSON.parse(lines[0]);
-    } catch (e) {
-      throw new Error(`Invalid JSONL format in ${filePath}`);
-    }
-    
-    return {
-      sessionId: firstMessage.sessionId,
-      timestamp: firstMessage.timestamp,
-      messageCount: lines.length,
-      gitBranch: firstMessage.gitBranch,
-      cwd: firstMessage.cwd,
-      fileSize: stats.size
-    };
-  }
-}
@@ -1,200 +0,0 @@
-import { existsSync, mkdirSync } from 'fs';
-import { join, dirname } from 'path';
-import { fileURLToPath } from 'url';
-import { HookError, CompressionError, Logger, FileLogger } from './types.js';
-
-const __filename = fileURLToPath(import.meta.url);
-const __dirname = dirname(__filename);
-
-export class ErrorHandler {
-  private logger: Logger;
-  private logDir: string;
-
-  // <Block> 7.1 ====================================
-  constructor(enableDebug = false) {
-    this.logDir = join(__dirname, '..', 'logs');
-    this.ensureLogDirectory();
-
-    const logFile = join(
-      this.logDir,
-      `claude-mem-${new Date().toISOString().slice(0, 10)}.log`
-    );
-    this.logger = new FileLogger(logFile, enableDebug);
-  }
-  // </Block> =======================================
-
-  // <Block> 7.2 ====================================
-  private ensureLogDirectory(): void {
-    if (!existsSync(this.logDir)) {
-      mkdirSync(this.logDir, { recursive: true });
-    }
-  }
-  // </Block> =======================================
-
-  // <Block> 7.3 ====================================
-  handleHookError(error: Error, hookType: string, payload?: unknown): never {
-    // <Block> 7.3a ====================================
-    const hookError =
-      error instanceof HookError
-        ? error
-        : new HookError(
-            error.message,
-            hookType,
-            payload as any,
-            'HOOK_EXECUTION_ERROR'
-          );
-    // </Block> =======================================
-
-    this.logger.error(`Hook execution failed in ${hookType}`, hookError, {
-      hookType,
-      payload: payload ? JSON.stringify(payload) : undefined,
-    });
-
-    console.log(
-      JSON.stringify({
-        continue: false,
-        stopReason: `Hook error: ${hookError.message}`,
-        error: {
-          type: hookError.name,
-          message: hookError.message,
-          code: hookError.code,
-        },
-      })
-    );
-
-    process.exit(1);
-  }
-  // </Block> =======================================
-
-  // <Block> 7.4 ====================================
-  handleCompressionError(
-    error: Error,
-    transcriptPath: string,
-    stage: string
-  ): never {
-    // <Block> 7.4a ====================================
-    const compressionError =
-      error instanceof CompressionError
-        ? error
-        : new CompressionError(error.message, transcriptPath, stage as any);
-    // </Block> =======================================
-
-    this.logger.error(`Compression failed during ${stage}`, compressionError, {
-      transcriptPath,
-      stage,
-    });
-
-    console.error(`Compression error: ${compressionError.message}`);
-    console.error(`Stage: ${stage}`);
-    console.error(`Transcript: ${transcriptPath}`);
-
-    process.exit(1);
-  }
-  // </Block> =======================================
-
-  // <Block> 7.5 ====================================
-  handleValidationError(
-    message: string,
-    context?: Record<string, unknown>
-  ): never {
-    this.logger.error('Validation error', undefined, { message, context });
-
-    console.error(`Validation error: ${message}`);
-    // <Block> 7.5a ====================================
-    if (context) {
-      console.error('Context:', JSON.stringify(context, null, 2));
-    }
-    // </Block> =======================================
-
-    process.exit(1);
-  }
-  // </Block> =======================================
-
-  // <Block> 7.6 ====================================
-  logSuccess(operation: string, details?: Record<string, unknown>): void {
-    this.logger.info(`Operation successful: ${operation}`, details);
-  }
-  // </Block> =======================================
-
-  // <Block> 7.7 ====================================
-  logWarning(message: string, details?: Record<string, unknown>): void {
-    this.logger.warn(message, details);
-  }
-  // </Block> =======================================
-
-  // <Block> 7.8 ====================================
-  logDebug(message: string, details?: Record<string, unknown>): void {
-    this.logger.debug(message, details);
-  }
-  // </Block> =======================================
-}
-
-// <Block> 7.9 ====================================
-export function parseStdinJson<T = unknown>(input: string): T {
-  try {
-    return JSON.parse(input) as T;
-  } catch (error) {
-    throw new Error(
-      `Failed to parse JSON input: ${error instanceof Error ? error.message : 'Unknown error'}`
-    );
-  }
-}
-// </Block> =======================================
-
-// <Block> 7.10 ===================================
-export async function safeExecute<T>(
-  operation: () => Promise<T>,
-  errorHandler: ErrorHandler,
-  context: string
-): Promise<T> {
-  try {
-    return await operation();
-  } catch (error) {
-    const message = `Safe execution failed in ${context}: ${error instanceof Error ? error.message : String(error)}`;
-    errorHandler.handleValidationError(message, { context, error });
-    throw error;
-  }
-}
-// </Block> =======================================
-
-// <Block> 7.11 ===================================
-export function validateFileExists(
-  filePath: string,
-  errorHandler: ErrorHandler
-): void {
-  if (!existsSync(filePath)) {
-    errorHandler.handleValidationError(`File not found: ${filePath}`, {
-      filePath,
-    });
-  }
-}
-// </Block> =======================================
-
-// <Block> 7.12 ===================================
-/**
- * Creates a standardized hook response using HookTemplates
- * @deprecated Use HookTemplates.createHookSuccessResponse or createHookErrorResponse instead
- * This function is maintained for backward compatibility but should be replaced with HookTemplates.
- */
-export function createHookResponse(
-  success: boolean,
-  data?: Record<string, unknown>
-): string {
-  // Log deprecation warning in development mode
-  if (process.env.NODE_ENV === 'development') {
-    console.warn('createHookResponse in error-handler.ts is deprecated. Use HookTemplates.createHookSuccessResponse or createHookErrorResponse instead.');
-  }
-
-  const response = {
-    continue: success,
-    suppressOutput: true, // Add standard suppressOutput field for Claude Code compatibility
-    ...data,
-  };
-
-  return JSON.stringify(response);
-}
-// </Block> =======================================
-
-export const globalErrorHandler = new ErrorHandler(
-  process.env.DEBUG === 'true'
-);
@@ -0,0 +1,42 @@
+import { appendFileSync, existsSync, mkdirSync } from 'fs';
+import { join } from 'path';
+import { PathDiscovery } from '../services/path-discovery.js';
+
+let logPath: string | null = null;
+
+function ensureLogPath(): string {
+  if (logPath) {
+    return logPath;
+  }
+
+  const discovery = PathDiscovery.getInstance();
+  const logsDir = discovery.getLogsDirectory();
+
+  if (!existsSync(logsDir)) {
+    mkdirSync(logsDir, { recursive: true });
+  }
+
+  logPath = join(logsDir, 'rolling-memory.log');
+  return logPath;
+}
+
+export type RollingLogLevel = 'debug' | 'info' | 'warn' | 'error';
+
+export function rollingLog(
+  level: RollingLogLevel,
+  message: string,
+  payload: Record<string, unknown> = {}
+): void {
+  try {
+    const file = ensureLogPath();
+    const entry = {
+      timestamp: new Date().toISOString(),
+      level,
+      message,
+      ...payload
+    };
+    appendFileSync(file, `${JSON.stringify(entry)}\n`, 'utf8');
+  } catch {
+    // Logging should never throw user-facing errors
+  }
+}
@@ -0,0 +1,87 @@
+import { readSettings } from './settings.js';
+
+export interface RollingSettings {
+  captureEnabled: boolean;
+  summaryEnabled: boolean;
+  sessionStartEnabled: boolean;
+  chunkTokenLimit: number;
+  chunkOverlapTokens: number;
+  summaryTurnLimit: number;
+}
+
+const DEFAULTS: RollingSettings = {
+  captureEnabled: true,
+  summaryEnabled: true,
+  sessionStartEnabled: true,
+  chunkTokenLimit: 600,
+  chunkOverlapTokens: 200,
+  summaryTurnLimit: 20
+};
+
+function normalizeBoolean(value: unknown, fallback: boolean): boolean {
+  if (typeof value === 'boolean') {
+    return value;
+  }
+  if (typeof value === 'string') {
+    const lowered = value.toLowerCase();
+    if (lowered === 'true') return true;
+    if (lowered === 'false') return false;
+  }
+  return fallback;
+}
+
+function normalizeNumber(value: unknown, fallback: number): number {
+  if (typeof value === 'number' && Number.isFinite(value)) {
+    return value;
+  }
+  if (typeof value === 'string') {
+    const parsed = Number(value);
+    if (!Number.isNaN(parsed) && Number.isFinite(parsed)) {
+      return parsed;
+    }
+  }
+  return fallback;
+}
+
+export function getRollingSettings(): RollingSettings {
+  const settings = readSettings();
+
+  return {
+    captureEnabled: normalizeBoolean(
+      settings.rollingCaptureEnabled,
+      DEFAULTS.captureEnabled
+    ),
+    summaryEnabled: normalizeBoolean(
+      settings.rollingSummaryEnabled,
+      DEFAULTS.summaryEnabled
+    ),
+    sessionStartEnabled: normalizeBoolean(
+      settings.rollingSessionStartEnabled,
+      DEFAULTS.sessionStartEnabled
+    ),
+    chunkTokenLimit: normalizeNumber(
+      settings.rollingChunkTokens,
+      DEFAULTS.chunkTokenLimit
+    ),
+    chunkOverlapTokens: normalizeNumber(
+      settings.rollingChunkOverlapTokens,
+      DEFAULTS.chunkOverlapTokens
+    ),
+    summaryTurnLimit: normalizeNumber(
+      settings.rollingSummaryTurnLimit,
+      DEFAULTS.summaryTurnLimit
+    )
+  };
+}
+
+export function isRollingCaptureEnabled(): boolean {
+  return getRollingSettings().captureEnabled;
+}
+
+export function isRollingSummaryEnabled(): boolean {
+  return getRollingSettings().summaryEnabled;
+}
+
+export function isRollingSessionStartEnabled(): boolean {
+  return getRollingSettings().sessionStartEnabled;
+}
@@ -1,103 +1,17 @@
-export interface HookPayload {
-  session_id: string;
-  transcript_path: string;
-  hook_event_name: string;
-}
+/**
+ * Core Type Definitions
+ *
+ * Minimal type definitions for the claude-mem system.
+ * Only includes types that are actively imported and used.
+ */

-export interface PreCompactPayload extends HookPayload {
-  hook_event_name: 'PreCompact';
-  trigger: 'manual' | 'auto';
-  custom_instructions?: string;
-}
-
-export interface SessionStartPayload extends HookPayload {
-  hook_event_name: 'SessionStart';
-  source: 'startup' | 'compact' | 'vscode' | 'web';
-}
-
-export interface UserPromptSubmitPayload extends HookPayload {
-  hook_event_name: 'UserPromptSubmit';
-  prompt: string;
-  cwd: string;
-}
-
-export interface PreToolUsePayload extends HookPayload {
-  hook_event_name: 'PreToolUse';
-  tool_name: string;
-  tool_input: Record<string, unknown>;
-}
-
-export interface PostToolUsePayload extends HookPayload {
-  hook_event_name: 'PostToolUse';
-  tool_name: string;
-  tool_input: Record<string, unknown>;
-  tool_response: Record<string, unknown> & {
-    success?: boolean;
-  };
-}
-
-export interface NotificationPayload extends HookPayload {
-  hook_event_name: 'Notification';
-  message: string;
-  title?: string;
-}
-
-export interface StopPayload extends HookPayload {
-  hook_event_name: 'Stop';
-  stop_hook_active: boolean;
-}
-
-export interface BaseHookResponse {
-  continue?: boolean;
-  stopReason?: string;
-  suppressOutput?: boolean;
-}
-
-export interface PreCompactResponse extends BaseHookResponse {
-  decision?: 'approve' | 'block';
-  reason?: string;
-}
-
-export interface SessionStartResponse extends BaseHookResponse {
-  hookSpecificOutput?: {
-    hookEventName: 'SessionStart';
-    additionalContext?: string;
-  };
-}
-
-export interface PreToolUseResponse extends BaseHookResponse {
-  permissionDecision?: 'allow' | 'deny' | 'ask';
-  permissionDecisionReason?: string;
-}
-
-export interface CompressionResult {
-  compressedLines: string[];
-  originalTokens: number;
-  compressedTokens: number;
-  compressionRatio: number;
-  memoryNodes: string[];
-}
-
-export interface MemoryNode {
-  id: string;
-  type: 'document';
-  content: string;
-  timestamp: string;
-  metadata?: Record<string, unknown>;
-}
-
-export class HookError extends Error {
-  constructor(
-    message: string,
-    public hookType: string,
-    public payload?: HookPayload,
-    public code?: string
-  ) {
-    super(message);
-    this.name = 'HookError';
-  }
-}
+// =============================================================================
+// ERROR CLASSES
+// =============================================================================

+/**
+ * Custom error class for compression failures
+ */
 export class CompressionError extends Error {
  constructor(
    message: string,
@@ -109,108 +23,8 @@ export class CompressionError extends Error {
  }
 }

-export interface Logger {
-  info(message: string, meta?: Record<string, unknown>): void;
-  warn(message: string, meta?: Record<string, unknown>): void;
-  error(message: string, error?: Error, meta?: Record<string, unknown>): void;
-  debug(message: string, meta?: Record<string, unknown>): void;
-}
-
-export class FileLogger implements Logger {
-  constructor(
-    private logFile: string,
-    private enableDebug = false
-  ) {}
-
-  info(message: string, meta?: Record<string, unknown>): void {
-    this.log('INFO', message, meta);
-  }
-
-  warn(message: string, meta?: Record<string, unknown>): void {
-    this.log('WARN', message, meta);
-  }
-
-  error(message: string, error?: Error, meta?: Record<string, unknown>): void {
-    const errorMeta = error ? { error: error.message, stack: error.stack } : {};
-    this.log('ERROR', message, { ...meta, ...errorMeta });
-  }
-
-  debug(message: string, meta?: Record<string, unknown>): void {
-    if (this.enableDebug) {
-      this.log('DEBUG', message, meta);
-    }
-  }
-
-  private log(
-    level: string,
-    message: string,
-    meta?: Record<string, unknown>
-  ): void {
-    const timestamp = new Date().toISOString();
-    const metaStr = meta ? ` ${JSON.stringify(meta)}` : '';
-    const logLine = `[${timestamp}] ${level}: ${message}${metaStr}\n`;
-
-    console.error(logLine);
-  }
-}
-
-export function validateHookPayload(
-  payload: unknown,
-  expectedType: string
-): HookPayload {
-  if (!payload || typeof payload !== 'object') {
-    throw new HookError(
-      `Invalid payload: expected object, got ${typeof payload}`,
-      expectedType
-    );
-  }
-
-  const hookPayload = payload as Record<string, unknown>;
-
-  if (!hookPayload.session_id || typeof hookPayload.session_id !== 'string') {
-    throw new HookError(
-      'Missing or invalid session_id',
-      expectedType,
-      hookPayload as unknown as HookPayload
-    );
-  }
-
-  if (
-    !hookPayload.transcript_path ||
-    typeof hookPayload.transcript_path !== 'string'
-  ) {
-    throw new HookError(
-      'Missing or invalid transcript_path',
-      expectedType,
-      hookPayload as unknown as HookPayload
-    );
-  }
-
-  return hookPayload as unknown as HookPayload;
-}
-
-export function createSuccessResponse(
-  additionalData?: Record<string, unknown>
-): BaseHookResponse {
-  return {
-    continue: true,
-    ...additionalData,
-  };
-}
-
-export function createErrorResponse(
-  reason: string,
-  additionalData?: Record<string, unknown>
-): BaseHookResponse {
-  return {
-    continue: false,
-    stopReason: reason,
-    ...additionalData,
-  };
-}
-
 // =============================================================================
-// SETTINGS AND CONFIGURATION TYPES
+// CONFIGURATION TYPES
 // =============================================================================

 /**
@@ -224,40 +38,11 @@ export interface Settings {
  embedded?: boolean;
  saveMemoriesOnClear?: boolean;
  claudePath?: string;
+  rollingCaptureEnabled?: boolean;
+  rollingSummaryEnabled?: boolean;
+  rollingSessionStartEnabled?: boolean;
+  rollingChunkTokens?: number;
+  rollingChunkOverlapTokens?: number;
+  rollingSummaryTurnLimit?: number;
  [key: string]: unknown;  // Allow additional properties
 }
-
-// =============================================================================
-// MCP CLIENT INTERFACE TYPES
-// =============================================================================
-
-/**
- * Document structure for MCP operations
- */
-export interface MCPDocument {
-  id: string;
-  content: string;
-  metadata?: Record<string, unknown>;
-}
-
-/**
- * Search result structure from MCP operations
- */
-export interface MCPSearchResult {
-  documents?: MCPDocument[];
-  ids?: string[];
-  metadatas?: Record<string, unknown>[];
-  distances?: number[];
-  [key: string]: unknown;
-}
-
-/**
- * Interface for MCP client implementations (Chroma-based)
- */
-export interface IMCPClient {
-  connect(): Promise<void>;
-  disconnect(): Promise<void>;
-  addDocuments(documents: MCPDocument[]): Promise<void>;
-  queryDocuments(query: string, limit?: number): Promise<MCPSearchResult>;
-  getDocuments(ids?: string[]): Promise<MCPSearchResult>;
-}
@@ -0,0 +1,112 @@
+import { platform, homedir } from 'os';
+import { execSync } from 'child_process';
+import { chmodSync } from 'fs';
+import { join } from 'path';
+
+const isWindows = platform() === 'win32';
+
+/**
+ * Platform-specific utilities for cross-platform compatibility
+ * Handles differences between Windows and Unix-like systems
+ */
+export const Platform = {
+  /**
+   * Returns the appropriate shell for the current platform
+   */
+  getShell: (): string => {
+    return isWindows ? 'powershell' : '/bin/sh';
+  },
+
+  /**
+   * Returns the file extension for hook scripts
+   */
+  getHookExtension: (): string => {
+    return '.js'; // Both platforms can execute Node.js scripts
+  },
+
+  /**
+   * Finds the path to an executable command
+   * @param name - Name of the executable to find
+   * @returns Full path to the executable
+   */
+  findExecutable: (name: string): string => {
+    const cmd = isWindows ? `where ${name}` : `which ${name}`;
+    return execSync(cmd, {
+      encoding: 'utf8',
+      stdio: ['ignore', 'pipe', 'ignore']
+    }).trim();
+  },
+
+  /**
+   * Makes a file executable (Unix only - no-op on Windows)
+   * @param path - Path to the file to make executable
+   */
+  makeExecutable: (path: string): void => {
+    if (!isWindows) {
+      chmodSync(path, 0o755);
+    }
+  },
+
+  /**
+   * Installs uv package manager using platform-specific method
+   */
+  installUv: (): void => {
+    if (isWindows) {
+      execSync('powershell -Command "irm https://astral.sh/uv/install.ps1 | iex"', {
+        stdio: 'pipe'
+      });
+    } else {
+      execSync('curl -LsSf https://astral.sh/uv/install.sh | sh', {
+        stdio: 'pipe',
+        shell: '/bin/sh'
+      });
+    }
+  },
+
+  /**
+   * Returns shell configuration file paths for the current platform
+   * @returns Array of shell config file paths
+   */
+  getShellConfigPaths: (): string[] => {
+    const home = homedir();
+
+    if (isWindows) {
+      return [
+        join(home, 'Documents', 'PowerShell', 'Microsoft.PowerShell_profile.ps1'),
+        join(home, 'Documents', 'WindowsPowerShell', 'Microsoft.PowerShell_profile.ps1')
+      ];
+    }
+
+    return [
+      join(home, '.bashrc'),
+      join(home, '.zshrc'),
+      join(home, '.bash_profile')
+    ];
+  },
+
+  /**
+   * Gets the appropriate alias syntax for the current platform's shell
+   * @param aliasName - Name of the alias
+   * @param command - Command to alias
+   * @returns Alias definition string
+   */
+  getAliasDefinition: (aliasName: string, command: string): string => {
+    if (isWindows) {
+      // PowerShell function syntax
+      return `function ${aliasName} { ${command} $args }`;
+    }
+
+    // Bash/Zsh alias syntax
+    return `alias ${aliasName}='${command}'`;
+  },
+
+  /**
+   * Returns whether the current platform is Windows
+   */
+  isWindows: (): boolean => isWindows,
+
+  /**
+   * Returns whether the current platform is Unix-like (macOS/Linux)
+   */
+  isUnix: (): boolean => !isWindows
+};
Author	SHA1	Message	Date
Alex Newman	5f15695c3f	Release v3.9.16 Published from npm package build Source: https://github.com/thedotmack/claude-mem-source	2025-10-06 22:55:16 -04:00
Alex Newman	c49533c250	Release v3.9.14 Published from npm package build Source: https://github.com/thedotmack/claude-mem-source	2025-10-03 21:47:35 -04:00
Alex Newman	4f49cb1bc9	Release v3.9.13 Published from npm package build Source: https://github.com/thedotmack/claude-mem-source	2025-10-03 21:45:36 -04:00
Alex Newman	874726b193	Release v3.9.12 Published from npm package build Source: https://github.com/thedotmack/claude-mem-source	2025-10-03 21:09:11 -04:00
Alex Newman	5244a12422	Release v3.9.11 Published from npm package build Source: https://github.com/thedotmack/claude-mem-source	2025-10-03 20:55:36 -04:00
Alex Newman	5b30764fa8	Release v3.9.10 Published from npm package build Source: https://github.com/thedotmack/claude-mem-source	2025-10-03 18:27:36 -04:00
Alex Newman	85ed7c3d2f	Release v3.9.9 Published from npm package build Source: https://github.com/thedotmack/claude-mem-source	2025-10-03 18:20:47 -04:00
Alex Newman	4d5b307a74	Release v3.7.2 Published from npm package build Source: https://github.com/thedotmack/claude-mem-source	2025-09-22 14:14:51 -04:00