Memory Persistence

Session lifecycle hooks that maintain context and knowledge across Claude Code sessions.

What This Does

Captures session metadata and context:

• Session Start: Initialize session directory, record metadata
• Session End: Generate summary, prepare for next session
• Session Logging: Track session activity for pattern extraction

Architecture

Session Start → Initialize → Capture Metadata
     │
     │ (work session)
     │
     ▼
Session End → Generate Summary → Extract Patterns
     │
     ▼
~/.claude/sessions/<session-id>/
├── start.jsonl    # Session start metadata
├── end.jsonl      # Session end metadata
├── session.log    # Activity log
└── summary.md     # Human-readable summary

Session Structure

Each session creates a directory in

~/.claude/sessions/

:

~/.claude/sessions/
├── 20250127-143022-12345/
│   ├── start.jsonl    # {"timestamp":"...","event":"session_start",...}
│   ├── end.jsonl      # {"timestamp":"...","event":"session_end",...}
│   ├── session.log    # Activity log
│   └── summary.md     # Session summary
├── 20250127-154530-12346/
│   ├── ...

Hooks Configuration

Configure in

.claude/settings.json

:

{
  "hooks": {
    "SessionStart": [
      {
        "matcher": ".*",
        "hooks": [
          {
            "type": "command",
            "command": "\"$CLAUDE_PROJECT_DIR\"/.claude/scripts/session-start.sh",
            "description": "Initialize session logging"
          }
        ]
      }
    ],
    "SessionEnd": [
      {
        "matcher": ".*",
        "hooks": [
          {
            "type": "command",
            "command": "\"$CLAUDE_PROJECT_DIR\"/.claude/scripts/session-end.sh",
            "description": "Finalize session summary"
          }
        ]
      }
    ]
  }
}

Session Metadata

Session Start Records:

• Timestamp
• Session ID (YYYYMMDD-HHMMSS-PID)
• Working directory
• Git branch
• Environment variables

Session End Records:

• Timestamp
• Session duration (calculated)
• Working directory
• Final state

Integration with Handoff

Memory persistence integrates with the

/handoff

command:

Session End → Handoff Generation:

1. Session ends (hook triggers)
2. Summary generated automatically
3. Handoff document created with session context
4. Next session loads handoff for continuity

Handoff Document Structure:

goal: [What was being worked on]
now: [Current state]
context:
  session_id: [Previous session ID]
  session_summary: [Link to session summary]
  files_modified: [List of files changed]
next_steps: [What to do next]

Cross-Platform Compatibility

Hooks use POSIX-compliant bash for cross-platform support:

• macOS: Works natively
• Linux: Works natively
• Windows: Works via Git Bash or WSL

Session Management

View Sessions:

# List all sessions
ls ~/.claude/sessions/

# View recent sessions
ls -lt ~/.claude/sessions/ | head -10

# Count total sessions
ls ~/.claude/sessions/ | wc -l

Clean Old Sessions:

# Remove sessions older than 30 days
find ~/.claude/sessions/ -type d -mtime +30 -exec rm -rf {} \;

Archive Important Sessions:

# Archive specific session
cp -r ~/.claude/sessions/20250127-143022-12345 ~/archive/sessions/

Best Practices

1. Review session summaries - Check summary.md before starting new session
2. Extract patterns - Use

   /learn

   to capture valuable patterns from sessions
3. Archive key sessions - Save important sessions for future reference
4. Clean periodically - Remove old sessions to manage disk space
5. Integrate with handoff - Use handoff for session continuity

Privacy Considerations

• Local storage only - Sessions stored on your machine
• No cloud sync - Session data never uploaded
• No conversation content - Only metadata and summaries
• User control - You decide when to archive or delete

Integration with Seed System

Handoff Command

Session hooks integrate with

/handoff

:

• Session end → Generate handoff document automatically
• Next session → Load previous handoff for continuity
• Seamless transition between sessions

Related Skills

• handoff - Session continuity and context transfer
• filesystem-context - Persistent storage patterns

Key Principle

Session lifecycle hooks maintain context continuity across sessions, enable pattern extraction, and integrate with handoff for seamless workflow transitions.

---

<critical_constraint> MANDATORY: Capture session metadata at start and end MANDATORY: Create handoff documents for session continuity MANDATORY: Never lose TaskList ID across sessions No exceptions. Session persistence enables multi-session workflows. </critical_constraint>

---

Genetic Code

This component carries essential Seed System principles for context: fork isolation:

<critical_constraint> MANDATORY: All components MUST be self-contained (zero .claude/rules dependency) MANDATORY: Achieve 80-95% autonomy (0-5 AskUserQuestion rounds per session) MANDATORY: Description MUST use What-When-Not format in third person MANDATORY: No component references another component by name in description MANDATORY: Progressive disclosure - references/ for detailed content MANDATORY: Use XML for control (mission_control, critical_constraint), Markdown for data No exceptions. Portability invariant must be maintained. </critical_constraint>

Delta Standard: Good Component = Expert Knowledge − What Claude Already Knows

Recognition Questions:

• "Would Claude know this without being told?" → Delete (zero delta)
• "Can this work standalone?" → Fix if no (non-self-sufficient)
• "Did I read the actual file, or just see it in grep?" → Verify before claiming

---