haddock-development/claude-reflect-system / reflect
reflect
VerifiedAnalyzes conversation transcripts to extract user corrections, patterns, and preferences, then proposes skill improvements. Use this skill when users provide corrections, express preferences about code style, or when patterns emerge from successful approaches. Can be triggered manually with /reflect or automatically at session end when enabled.
| File | Kind | Size |
|---|---|---|
| .state/auto-reflection.json | config | 54 B |
| .state/last-reflection.timestamp | other | 0 B |
| README.md | markdown | 7.3 KB |
| SKILL.md | skill md | 2.8 KB |
| USER_GUIDE.md | markdown | 21.4 KB |
| commands/reflect-meta.md | markdown | 1.6 KB |
| commands/reflect-promote.md | markdown | 1.8 KB |
| commands/reflect-stats.md | markdown | 772 B |
| references/signal-patterns.md | markdown | 6.0 KB |
| scripts/extract_signals.pyexecutable | script | 12.3 KB |
| scripts/hook-stop.shexecutable | script | 2.2 KB |
| scripts/learning_ledger.py | script | 12.6 KB |
| scripts/meta_learning.py | script | 14.8 KB |
| scripts/present_review.pyexecutable | script | 9.9 KB |
| scripts/promote_learning.py | script | 8.0 KB |
| scripts/reflect.pyexecutable | script | 5.3 KB |
| scripts/scope_analyzer.py | script | 8.3 KB |
| scripts/semantic_detector.py | script | 8.1 KB |
| scripts/toggle-off.shexecutable | script | 431 B |
| scripts/toggle-on.shexecutable | script | 499 B |
| scripts/toggle-status.shexecutable | script | 2.5 KB |
| scripts/update_skill.pyexecutable | script | 8.2 KB |
commands/reflect-promote.md
View on GitHub---
description: Promote learnings from skills to global CLAUDE.md
allowed-tools: Bash, Read, Write
---
# /reflect-promote
Promote skill-level learnings to global scope when they appear in multiple repositories.
## What This Does
When you correct Claude in multiple projects, those learnings become candidates for promotion to your global `~/.claude/CLAUDE.md` file. This means Claude will remember them everywhere, not just in specific skills.
## Usage
```bash
# List learnings ready for promotion
python3 ~/.claude/skills/reflect/scripts/promote_learning.py list
# Preview what would be added
python3 ~/.claude/skills/reflect/scripts/promote_learning.py preview <fingerprint>
# Promote a specific learning
python3 ~/.claude/skills/reflect/scripts/promote_learning.py promote <fingerprint>
# Promote all eligible learnings
python3 ~/.claude/skills/reflect/scripts/promote_learning.py all
# Dry-run (preview without changes)
python3 ~/.claude/skills/reflect/scripts/promote_learning.py all --dry-run
```
## How It Works
1. **Track**: Every `/reflect` session records learnings in SQLite ledger
2. **Detect**: When same learning appears in 2+ repos → eligible for promotion
3. **Promote**: Append to `~/.claude/CLAUDE.md` with metadata
4. **Backup**: Original file backed up before changes
## Example
```
$ python3 promote_learning.py list
2 learnings ready for promotion:
[a1b2c3d4] (3 repos)
Use uv instead of pip for Python projects
From: python-project-creator
[e5f6g7h8] (2 repos)
Always run tests before committing
From: general
```
## Thresholds
- Default: 2 repos required for promotion
- Can be configured in scope_analyzer.py
## Files
- Ledger DB: `~/.claude/reflect/learnings.db`
- Global rules: `~/.claude/CLAUDE.md`
- Backups: `~/.claude/backups/`
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |