From 5c4b44d4c7db845597fad540f172572f76dcf9b7 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Fri, 31 Oct 2025 21:02:02 +0000
Subject: [PATCH 1/3] Initial plan


From 140a5f346993438905010ae79ad44fbe38526f41 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Fri, 31 Oct 2025 21:06:59 +0000
Subject: [PATCH 2/3] Clarify Stop hook behavior and /clear vs exit in docs

- Changed "Stop Claude" to "Claude finishes responding" to clarify that the Stop hook fires automatically
- Explained Stop hook triggers summary generation automatically, not as a user action
- Added detailed section explaining /clear behavior vs completely exiting Claude Code
- Clarified when new context is injected (new session start, not after /clear)
- Fixed session count: changed "last 3 sessions" to "last 10 sessions" (matches implementation)

Co-authored-by: thedotmack <683968+thedotmack@users.noreply.github.com>
---
 docs/usage/getting-started.mdx | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

diff --git a/docs/usage/getting-started.mdx b/docs/usage/getting-started.mdx
index 4f36e9ba..d2c2ae3a 100644
--- a/docs/usage/getting-started.mdx
+++ b/docs/usage/getting-started.mdx
@@ -11,9 +11,9 @@ Claude-Mem works automatically once installed. No manual intervention required!
 
 ### The Full Cycle
 
-1. **Start Claude Code** - Context from last 3 sessions appears automatically
+1. **Start Claude Code** - Context from last 10 sessions appears automatically
 2. **Work normally** - Every tool execution is captured
-3. **Stop Claude** - Summary is generated and saved
+3. **Claude finishes responding** - Stop hook automatically generates and saves a summary
 4. **Next session** - Previous work appears in context
 
 ### What Gets Captured
@@ -42,7 +42,7 @@ The worker service processes tool observations and extracts:
 
 ### Session Summaries
 
-When you stop Claude (or a session ends), a summary is generated with:
+When Claude finishes responding (triggering the Stop hook), a summary is automatically generated with:
 
 - **Request** - What you asked for
 - **Investigated** - What Claude explored
@@ -159,7 +159,7 @@ Context injection uses three-tier verbosity for efficient token usage:
 
 This ensures you get maximum detail for recent work while still having context from older sessions.
 
-## Multi-Prompt Sessions
+## Multi-Prompt Sessions & `/clear` Behavior
 
 Claude-Mem supports sessions that span multiple user prompts:
 
@@ -167,7 +167,15 @@ Claude-Mem supports sessions that span multiple user prompts:
 - **prompt_number**: Identifies specific prompt within session
 - **Session continuity**: Observations and summaries link across prompts
 
-When you use `/clear`, the session doesn't end - it continues with a new prompt number. This preserves context across conversation restarts.
+### Important Note About `/clear`
+
+When you use `/clear`, the session doesn't end - it continues with a new prompt number. This means:
+
+- ✅ **Observations are still being captured** and added to the current session
+- ✅ **A summary will be generated** when Claude finishes responding (Stop hook fires)
+- ❌ **New context won't be injected** until you start a completely new session (exit and restart Claude Code)
+
+To get fresh context from previous sessions, you need to **exit Claude Code completely** (Ctrl+C Ctrl+C or close the terminal) and start a new session. The `/clear` command only clears the conversation context visible to Claude, but the underlying session continues.
 
 ## Next Steps
 

From 192adf5cbc08a033c65c34c9508db9a44364dec6 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Fri, 31 Oct 2025 21:28:54 +0000
Subject: [PATCH 3/3] Fix: /clear DOES inject context via SessionStart hook

Corrected incorrect documentation that claimed /clear doesn't inject context.
The SessionStart hook fires with source: "clear" when /clear is used, which
re-injects context from recent sessions. Updated docs to accurately reflect
that /clear both clears conversation AND re-injects fresh context.

Co-authored-by: thedotmack <683968+thedotmack@users.noreply.github.com>
---
 docs/usage/getting-started.mdx | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/usage/getting-started.mdx b/docs/usage/getting-started.mdx
index d2c2ae3a..20b516ea 100644
--- a/docs/usage/getting-started.mdx
+++ b/docs/usage/getting-started.mdx
@@ -171,11 +171,11 @@ Claude-Mem supports sessions that span multiple user prompts:
 
 When you use `/clear`, the session doesn't end - it continues with a new prompt number. This means:
 
+- ✅ **Context is re-injected** from recent sessions (SessionStart hook fires with `source: "clear"`)
 - ✅ **Observations are still being captured** and added to the current session
 - ✅ **A summary will be generated** when Claude finishes responding (Stop hook fires)
-- ❌ **New context won't be injected** until you start a completely new session (exit and restart Claude Code)
 
-To get fresh context from previous sessions, you need to **exit Claude Code completely** (Ctrl+C Ctrl+C or close the terminal) and start a new session. The `/clear` command only clears the conversation context visible to Claude, but the underlying session continues.
+The `/clear` command clears the conversation context visible to Claude AND re-injects fresh context from recent sessions, while the underlying session continues tracking observations.
 
 ## Next Steps