- Updated the formatSearchTips function to emphasize the importance of using index format first for token efficiency.
- Revised descriptions for various search tools to include reminders about starting with index format and using full format only for specific items of interest.
- Clarified output format descriptions to recommend index format for initial searches.
Changed summary prompts to generate discrete per-request summaries instead of cumulative session summaries. This provides better chronological memory where each summary is a clean unit representing one request/response cycle.
Changes:
- Renamed buildFinalizePrompt() to buildSummaryPrompt() in src/sdk/prompts.ts
- Updated prompt text to focus on "THIS REQUEST" rather than "this session"
- Updated all import and function call sites in worker-service.ts and worker.ts
- Added IMPORTANT warning to emphasize request-level scope
Expected behavior: Each summary will now describe only what happened during that specific request, eliminating cumulative recaps.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Updated contextHook to support colorized output for terminal and JSON format for hooks.
- Introduced ANSI color codes for improved readability in terminal output.
- Modified the output structure to include session details with color formatting.
- Added a new method in SessionStore to aggregate files read and modified from observations for a session.
- Improved error handling for JSON parsing of file data in the new method.
- Introduced `formatSearchTips` function to provide helpful search tips in the output.
- Added formatting options for search results, allowing users to choose between 'index' (titles/dates only) and 'full' (complete details).
- Updated handlers for observation and session search tools to accommodate the new format option.
- Enhanced result formatting to include date information and improved output structure.
### Changed
- Graceful session cleanup: Cleanup hook now marks sessions as completed instead of sending DELETE requests to worker
- Natural worker shutdown: Workers now finish pending operations (like summary generation) before terminating
- Restored MCP search server: Re-enabled full-text search capabilities from backup
### Fixed
- Session summaries no longer interrupted by aggressive cleanup during session end
- Workers can now complete final operations before shutdown
### Dependencies
- Updated @anthropic-ai/claude-agent-sdk to ^0.1.23
- Added @modelcontextprotocol/sdk ^1.20.1
- Added zod-to-json-schema ^3.24.6
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Replaced `createSdkMcpServer` with `Server` from `@modelcontextprotocol/sdk/server/index.js`.
- Updated tool definitions to use structured input schemas with Zod.
- Enhanced response formatting for search results, combining multiple results into a single text response.
- Added new tools for advanced search and recent session context retrieval.
- Improved error handling and logging throughout the server.
Simplified dependency installation by moving from TypeScript runtime bootstrap to bash-based checks in plugin manifest. This reduces complexity and code size while maintaining the same functionality.
Changes:
- Added bash conditional dependency checks to all 5 hooks in hooks.json
- Check runs before each hook: [ ! -d "${CLAUDE_PLUGIN_ROOT}/scripts/node_modules" ] && cd "${CLAUDE_PLUGIN_ROOT}/scripts" && npm install || true
- Reverted all hook TypeScript files to use simple static imports (removed dynamic imports)
- Removed src/shared/bootstrap.ts (44 lines)
- Removed ensureDependencies() calls from all hook entry points
Benefits:
- Simpler architecture using native bash instead of TypeScript
- Net reduction of 157 lines of code
- No runtime overhead when dependencies already installed
- Uses plugin manifest's command hook feature as intended
- Faster and more efficient
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Changed all hook entry points to use dynamic imports after bootstrap runs. This ensures that better-sqlite3 is installed before Node.js attempts to resolve the import.
Changes:
- Modified src/bin/hooks/*.ts to call ensureDependencies() before dynamic import
- Moved from static `import { hook } from '...'` to `const { hook } = await import('...')`
- This delays module resolution until after npm install completes
- Bumped version to 4.0.6
The previous approach failed because static imports are resolved at module link time, before any runtime code (including ensureDependencies) executes.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
The source field must be a relative path starting with ./ not a GitHub URL.
GitHub marketplace installation uses shorthand format:
/plugin marketplace add thedotmack/claude-mem
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Fixed better-sqlite3 distribution by implementing self-bootstrapping hooks
that auto-install dependencies on first run. This eliminates the need for
users to have native compilation tools or manually install dependencies.
## Solution
Instead of bundling 25MB of better-sqlite3 binaries in git or requiring
manual npm install, hooks now bootstrap themselves on first execution:
1. Created `src/shared/bootstrap.ts` with `ensureDependencies()` function
2. Added bootstrap calls to all hook entry points (context, new, save, summary, cleanup)
3. Created `plugin/scripts/package.json` declaring better-sqlite3 dependency
4. Bootstrap checks if `node_modules` exists, runs `npm install` if missing
5. npm automatically downloads prebuilt better-sqlite3 binary for user's platform
## Changes
**Core Bootstrap System:**
- Added src/shared/bootstrap.ts: Auto-install dependencies using npm
- Modified all hooks (context, new, save, summary, cleanup) to call ensureDependencies()
- Created plugin/scripts/package.json with better-sqlite3 dependency
**Build & Distribution:**
- Removed node_modules copying logic from build script
- Build output is now compact (hooks + package.json, no binaries)
- Updated marketplace.json to point to GitHub for direct installation
**Documentation:**
- Updated README: GitHub Marketplace installation is now recommended method
- Installation instructions emphasize no compilation needed
- Version bumped to 4.0.5 throughout
## Benefits
- ✅ No git bloat (repo stays small, no 25MB binaries committed)
- ✅ No compilation needed (npm downloads prebuilt binaries)
- ✅ Works on all platforms (npm handles platform-specific binaries)
- ✅ Zero manual steps (hooks bootstrap themselves automatically)
- ✅ Idempotent (skips install if dependencies already exist)
Installation now works via simple:
```
/plugin marketplace add https://github.com/thedotmack/claude-mem
/plugin install claude-mem
```
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Reverted from GitHub-hosted marketplace to local marketplace file installation method. This allows us to resolve issues with better-sqlite3 native module builds before enabling GitHub marketplace distribution.
Changes:
- Simplified .claude-plugin/marketplace.json (removed metadata, version fields)
- Updated README installation instructions to use local .claude-plugin/marketplace.json
- Bumped version to 4.0.4
Installation now requires cloning the repo and using:
/plugin marketplace add .claude-plugin/marketplace.json
Will restore GitHub marketplace method once native module issues are resolved.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
All hooks now call ensureWorkerRunning() BEFORE querying the database. This
ensures the worker's orphaned session cleanup runs before hooks check for
active sessions, preventing 404 errors when hooks try to use sessions that
don't exist in worker memory after a restart.
Hook order now:
1. ensureWorkerRunning() - starts worker, runs cleanup
2. Query DB - cleanup already marked orphaned sessions as failed
3. Use session - only valid sessions are processed
Fixed in:
- new-hook: Line 26, before DB queries
- save-hook: Line 37, before DB queries
- summary-hook: Line 24, before DB queries
- cleanup-hook: Line 50, before DB queries
This prevents the race condition where hooks would read session status before
cleanup ran, then get 404 from worker after cleanup marked sessions failed.
Updates marketplace.json with:
- Marketplace metadata (description, version)
- Plugin version synced to 4.0.2
- Full plugin description from plugin.json
- Author information for discoverability
Users can now install via:
/plugin marketplace add thedotmack/claude-mem
/plugin install claude-mem
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Updated `new-hook.js` to streamline logging and session management.
- Enhanced database schema initialization and migration processes.
- Improved error handling for worker startup in `worker-utils.ts`, ensuring PM2 binary and ecosystem config existence before spawning.
- Added detailed error messages for missing dependencies and spawn failures.
- Removed try-catch blocks in new-hook, save-hook, and summary-hook for cleaner flow.
- Enhanced error handling in save and summary hooks to throw errors instead of logging and returning.
- Introduced ensureWorkerRunning utility to manage worker service lifecycle and health checks.
- Replaced dynamic port allocation with a fixed port for the worker service.
- Simplified path management and removed unused port allocator utility.
- Added database schema initialization for fresh installations and improved migration handling.
BREAKING CHANGES:
- Data directory moved from ~/.claude-mem/ to ${CLAUDE_PLUGIN_ROOT}/data/
- Fresh start required - no migration from v3.x databases
- Worker service now auto-starts on SessionStart hook
New Features:
- MCP Search Server with 6 specialized search tools
- FTS5 full-text search across observations and sessions
- Auto-starting worker service in SessionStart hook
- Citation support for search results (claude-mem:// URIs)
Changes:
- Updated paths.ts to use CLAUDE_PLUGIN_ROOT for data directory
- Added worker auto-start logic to context hook
- Updated worker service to write port file to plugin data dir
- Bumped version to 4.0.0 in package.json and plugin.json
- Created comprehensive CHANGELOG.md documenting v4.0.0 changes
- Updated README.md with v4.0.0 breaking changes and features
- Rebuilt all hooks and worker service
Technical Improvements:
- Improved error handling and graceful degradation
- Structured logging across all components
- Enhanced plugin integration with Claude Code
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Comprehensive README update to align with current feature set:
NEW FEATURES:
- Documented MCP Search Server with 6 specialized tools
- search_observations: Full-text search across observations
- search_sessions: Full-text search across sessions
- find_by_concept: Find by concept tags
- find_by_file: Find by file references
- find_by_type: Find by observation type
- advanced_search: Combined search with filters
- Added usage examples for MCP tools with sample queries
- Documented FTS5 full-text search capabilities
- Added MCP server configuration section
ARCHITECTURE UPDATES:
- Added src/servers/ directory with search-server.ts
- Added SessionSearch.ts to database layer
- Added .mcp.json to plugin configuration
- Updated directory structure to show all built executables
- Added Search Pipeline data flow diagram
- Documented claude-mem:// citation URI scheme
FIXES DOCUMENTED:
- Context hook now uses proper hookSpecificOutput JSON format
- Updated SessionStart hook documentation
Build process updated to include search server compilation.
Changelog updated with v3.9.17 changes.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
The context hook was not appearing in Claude Code sessions because it was
outputting plain text to stdout instead of using the required JSON structure
for SessionStart hooks.
Changes:
- src/hooks/context.ts: Changed contextHook to return string instead of void,
removing direct console.log calls to make it more reusable
- src/bin/hooks/context-hook.ts: Wrap contextHook output in hookSpecificOutput
JSON structure with hookEventName "SessionStart" and additionalContext field
- Both TTY and stdin code paths now properly format and exit with code 0
Fixes the issue where recent session context was not being injected at session
start. Tested with npm run test:context - hook now properly outputs JSON with
recent sessions formatted as markdown.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Added search functionality for observations and sessions using full-text search.
- Implemented formatting functions for search results with citations.
- Created multiple tools for searching by various criteria including concept, file, type, and advanced search.
- Integrated structured filters and pagination options for search queries.
- Established error handling for search operations and server initialization.
- Added SDK session, observation, and summary types to types.ts.
- Refactored worker-service to use SessionStore for session management.
- Created SessionSearch class for FTS5 full-text search and structured queries.
- Implemented SessionStore for CRUD operations on SDK sessions, observations, and summaries.
- Added migrations for database schema updates, including new columns and constraints.
- Enhanced search capabilities with filters for projects, types, concepts, and date ranges.
- Updated SQL queries in summary-hook.js and HooksDatabase.ts to improve performance and clarity.
- Introduced a subquery for fetching recent sessions with status, ensuring correct ordering and grouping.
- Enhanced logging functionality for better debugging and error tracking.
- Added comprehensive documentation in README.md to outline the new features and usage instructions.
- Updated `summary-hook.js` to improve logging and session handling.
- Modified `context.ts` to fetch recent sessions with status and summary info, enhancing output formatting.
- Added new methods in `HooksDatabase.ts` for retrieving recent sessions and their summaries.
- Improved observation retrieval logic in `context.ts` to display relevant information for active sessions.
- Enhanced prompt documentation in `prompts.ts` to clarify output expectations.
- Refactored logger methods in `logger.ts` to instance methods for better encapsulation.
- Introduced a new Logger utility to standardize logging with correlation IDs and structured context.
- Replaced console.error and console.log statements with logger methods in various modules including save.ts, summary.ts, parser.ts, HooksDatabase.ts, and worker-service.ts.
- Enhanced error handling and logging for better traceability of observations and summaries.
- Made observations.text nullable in the database schema to support structured fields.
- Added correlation IDs for tracking observations through the processing pipeline.
All 4 hook entry point scripts were missing process.exit(0) after successful
execution, causing Node processes to hang indefinitely instead of returning
control to Claude Code with exit code 0.
Root cause: In commit 6f62a56, process.exit(0) calls were removed from the
hook functions but were never added to the entry point scripts that wrap them.
Fixed files:
- src/bin/hooks/save-hook.ts (PostToolUse)
- src/bin/hooks/new-hook.ts (UserPromptSubmit)
- src/bin/hooks/summary-hook.ts (Stop)
- src/bin/hooks/context-hook.ts (SessionStart)
This restores proper hook exit behavior and prevents Claude Code from waiting
indefinitely for hook completion.
- Updated observation schema to include hierarchical fields: title, subtitle, facts, narrative, concepts, files_read, and files_modified.
- Modified the save-hook and summary-hook scripts to accommodate the new observation structure.
- Added migration logic to the HooksDatabase for adding new fields to the observations table.
- Refactored the parser to extract new fields from XML formatted observations.
- Adjusted prompt generation to reflect the new observation format and requirements.
- Updated worker service to handle new observation and summary structures.
- Added final finalize prompt for session summary generation with required XML fields.
- Introduced recommended prompt flow with structured observation format and hierarchical storage principles.
- Created final init prompt for processing tool executions with clear guidelines on when to store observations.
- Developed final observation prompt for analyzing tool outputs and generating structured observations.
- Migrated old prompt flow to a new system with improved clarity and structured data handling.
- Updated parser and storage mechanisms to accommodate new observation formats and fields.
- Enhanced documentation for new prompts and their usage in memory processing sessions.
- Introduced a new hierarchical memory format for observations, including title, subtitle, facts, narrative, concepts, and files.
- Updated session start, tool execution, and session end prompts to reflect new structure and guidance.
- Replaced bash command execution with XML parsing for observation storage, improving reliability and reducing complexity.
- Established clear criteria for what to store and skip, eliminating ambiguous language and tool-type bias.
- Enhanced database schema to support new observation fields and relationships, ensuring data integrity.
- Added comprehensive session summaries at the end of each session, capturing key insights and next steps.
- Improved retrieval patterns for observations, allowing for granular searches by concept and file.
- Outlined future enhancements for semantic search and cross-session memory linking.
Alex Newman