| name | reflect |
| description | Analyze recent conversation to identify improvements for CLAUDE.md memory files. Trigger when user says "reflect on this session", "reflect on this conversation", "reflect on this code", or "improve CLAUDE.md by reflecting on this session". The verb "reflect" combined with a session/conversation/code subject is the key trigger. |
CLAUDE.md Reflection
Analyze recent conversation history to identify improvements for CLAUDE.md memory files.
Workflow
Phase 1: Analysis
Find all CLAUDE.md files:
python scripts/find_claude_md.py
Analyze conversation for:
- Repeated corrections from the user
- Misunderstandings of user requests
- Missing context that caused confusion
- Inconsistencies between responses and user expectations
- Instructions that belong in a different memory file
Read reference materials:
references/memory-locations.md- Memory file hierarchy and placementreferences/anti-patterns.md- Common mistakes to avoid
Phase 2: Interaction
Present findings using AskUserQuestion with checkboxes.
For each suggestion:
- Explain the issue found in conversation
- Propose specific change (add, update, or delete)
- Identify which memory file to update
- Show the exact text to add/modify
Example:
Issue: Repeatedly used 'var' instead of 'const' in JavaScript
Proposal: Add "NEVER use var in JavaScript, use const or let"
File: ./frontend/CLAUDE.md
Wait for user approval before implementing.
Phase 3: Implementation
For approved changes:
Select correct memory file:
./CLAUDE.md- Project-wide, shared with team./.claude/CLAUDE.md- Alternative project location./.claude/rules/*.md- Modular topic-specific rules~/.claude/CLAUDE.md- Personal preferences (all projects)./CLAUDE.local.md- Personal project-specific (gitignored)
Apply changes:
- Keep minimal and targeted
- Use bullet points, not paragraphs
- Be specific ("Use 2-space indentation" not "Format code properly")
- Add emphasis (NEVER, ALWAYS) only where critical
- Maintain existing structure
Avoid anti-patterns:
- NO hyperbolic language (crucial, proper, robust)
- NO vague instructions (write good code, use best practices)
- NO historical comments (we used to do X)
- NO redundant information
Memory File Hierarchy
| Type | Location | Purpose | Shared |
|---|---|---|---|
| Project | ./CLAUDE.md |
Team instructions | Yes (git) |
| Rules | ./.claude/rules/*.md |
Modular topic rules | Yes (git) |
| User | ~/.claude/CLAUDE.md |
Personal preferences | No |
| Local | ./CLAUDE.local.md |
Personal project prefs | No (gitignored) |
Files higher in hierarchy take precedence. Use @path/to/file syntax to import files.
Resources
scripts/find_claude_md.py- Locate all CLAUDE.md filesreferences/memory-locations.md- Detailed memory hierarchy docsreferences/anti-patterns.md- Common mistakes to avoid