evolve
Skill2 starsChanged 9 months agoMITView on GitHub
Auto-maintain Claude Code config files as your codebase evolves
41 other projects use a version of this.
Adopted by
Most recently taken up by
- SethGammon/Citadeltheir own version, first seen Nov 4, 2026
- vidit19sharma/neogentstheir own version, first seen Nov 1, 2026
- cabernect/caberostheir own version, first seen Nov 1, 2026
- Snowflake-Labs/coco-skillstheir own version, first seen Nov 1, 2026
- NobiResearch/Ovenbirdtheir own version, first seen Nov 1, 2026
- ES-92/esctheir own version, first seen Oct 20, 2026
Used with
Projects with it also use instinct-export, instinct-import and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
TheKREATORSSkillinstinct-export
Export instincts from project/global scope to a file
Also in 4 projects that use this
TheKREATORSSkillinstinct-import
Import instincts from file or URL into project/global scope
Also in 4 projects that use this
AWSSkillmap
Map verified service capabilities against customer-provided control framework and threat model. Produces mapping-results.json, controls-catalog.md, and framework-mapping.md. Supports pluggable frameworks: CCMv4, NIST 800-53, CIS Benchmarks, ISO 27001, etc. Split into 4 focused sub-skills for better accuracy; this skill orchestrates them.
Also in 5 projects that use this
diego-tobalinaSkillbuild-fix
Fix build and compilation errors systematically. Run when build fails.
Also in 3 projects that use this
launch52-aiSkillskill-create
Create and validate Claude Code skills with proper structure, checklists, and validation scripts. Use when creating a new skill, auditing existing skills, or ensuring skills meet quality standards. Generates complete skill scaffolds with SKILL.md, check scripts, guides, and templates.
Also in 3 projects that use this
qtzx06Skilleval
Evaluate trained YOLO model accuracy with mAP metrics and per-class error analysis. Identifies failure cases for targeted improvement. Use after training.
Also in 5 projects that use this
- +1 more from TheKREATORS
Other versions41 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 41
- revfactory/harness★ 9.1ksame name
- Q00/ouroboros★ 6.3ksame name
- yologdev/yoyo-evolve★ 1.9ksame name
- iovisor/ubpf★ 996same name
- SethGammon/Citadel★ 927same name
- GAIR-NLP/ASI-Evolve★ 867same name
- Kanevry/session-orchestrator★ 53same name
- mycelium-hq/ai-brain-starter★ 50same name
- Snowflake-Labs/coco-skills★ 49same name
- unclejobs-ai/second-claude-code★ 41same name
31 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository12 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- add-core-guideSkill
- add-project-guideSkill
- bootstrapSkill
- refactor-docsSkill
- releaseSkill
- researchSkill
- setupSkill
- sync-skillsSkill
- update-readmeSkill
- verify-docsSkill
- setupSkill
The file
We split this skill into the rules it gives an agent
10 rules. 0 are common, 4 are distinctive, and 6 are very specific to this project.
0
Common
Most of the 36 edited versions we read keep these.
4
Distinctive
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 19
Check whether the Stop hook is registered, and ask the user before installing:
- Very specific: line 21
Read .claude/settings.json (if missing or empty, start with {})
- Very specific: line 22
Check if any existing Stop hook prompt already contains "evolve" or "config maintenance"
- Very specific: line 33
If hooks key is missing, create {"hooks": {"Stop": [{"hooks": [<entry>]}]}}
- Very specific: line 34
If hooks.Stop exists, append the new hook entry to the existing array's hooks list
- Distinctive: line 52
No meaningful changes detected
- Very specific: line 53
ALL changed files are under .claude/ (avoid self-referential loops)
- Distinctive: line 69
Only flag files where the content is actually inconsistent with the current codebase state.
- Distinctive: line 76
New skills: create only if a clear repeatable pattern was found (at least 2 similar operations in the current session)
- Very specific: line 77
CHANGELOG: add a concise one-line entry under the appropriate subsection in [Unreleased]
“Common” is measured: kept by at least 6 in 10 of the 36 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
/evolve — Auto-Maintain Claude Code Configuration
Automatically keep Claude Code config files accurate as a codebase evolves. Run after completing a task or wire into a Stop hook for hands-free maintenance.
Phase 0 — Self-Install (with consent)
Check whether the Stop hook is registered, and ask the user before installing:
- Read
.claude/settings.json(if missing or empty, start with{}) - Check if any existing Stop hook prompt already contains
"evolve"or"config maintenance" - If already present: continue silently to Phase 1
- If not found: show the user the hook JSON that would be added:
{ "type": "prompt", "prompt": "Before finishing, silently evaluate if any Claude Code configuration files need updating based on the work just completed. Check CLAUDE.md accuracy, CHANGELOG.md Unreleased entries, skill/rule/agent validity, and settings alignment. Only edit files that are actually stale. Report changes in one line or stay silent." } - Ask the user: "Install Stop hook for automatic config maintenance? (y/n)"
- If confirmed: merge the hook into settings and save:
- If
hookskey is missing, create{"hooks": {"Stop": [{"hooks": [<entry>]}]}} - If
hooks.Stopexists, append the new hook entry to the existing array'shookslist - Preserve all other keys in the settings file
- Report:
[evolve] Installed: Stop hook added for automatic config maintenance
- If
- If declined: report
[evolve] Skipped: Stop hook not installed. Run /evolve manually when needed.and continue to Phase 1
Phase 1 — Detect What Changed
Gather recent changes to understand what work was done:
git diff --name-only
git diff --staged --name-only
git log --oneline -5Categorize every changed file into: source code, tests, configs, dependencies, docs, new files/dirs.
Exit silently if:
- No meaningful changes detected
- ALL changed files are under
.claude/(avoid self-referential loops)
Phase 2 — Evaluate Each Config File
For each Claude Code config file that exists in the project, check whether it is stale:
| Config File | What to Check |
CLAUDE.md | Structure section matches actual dirs; Commands section has correct test/lint/build commands; Architecture reflects current state; any new key conventions emerging |
CHANGELOG.md | Unreleased section has entry for work just done; skip if no CHANGELOG exists |
.claude/skills/*/SKILL.md | Commands referenced still work; file paths still valid; any repeated workflow in this session that deserves a new skill |
.claude/rules/*.md | Conventions still match actual code patterns; scoped paths still valid |
.claude/agents/*.md | Tool lists still valid; model choices still appropriate |
.claude/settings.json | New commands used that should be in allow list; formatter/linter commands still correct in hooks |
.mcp.json | Existing config still valid (don't auto-add servers) |
Only flag files where the content is actually inconsistent with the current codebase state.
Phase 3 — Apply Minimal Updates
- Only edit files that are actually stale — skip everything else
- Use targeted edits — never rewrite an entire file
- Respect existing style — match the formatting of the file being edited
- New skills: create only if a clear repeatable pattern was found (at least 2 similar operations in the current session)
- CHANGELOG: add a concise one-line entry under the appropriate subsection in
[Unreleased]
Phase 4 — Report
- Changes made:
[evolve] Updated: CLAUDE.md (new dir documented), CHANGELOG.md (added feature entry) - New skill suggested:
[evolve] Suggestion: consider creating /deploy skill for the deployment workflow used today - Nothing needed: complete silence — no output at all
Rules
- NEVER commit — only edit files, let the user decide when to commit
- NEVER create files without clear justification — threshold: repeatable pattern seen at least 2 times in the session
- Skip if only
.claude/files changed — avoids self-referential loops - Conservative — when uncertain, suggest instead of edit
- Fast — no sub-skill invocations, no web fetches, just local file reads and edits
- Respect existing style — match formatting of the file being edited
- Project-agnostic — no hardcoded paths, languages, or tools; all detection-based
Stop Hook
The /evolve skill offers to install its Stop hook on first run (Phase 0). You will be asked for consent before any settings are modified.
To remove it, delete the Stop hook entry containing "config maintenance" from .claude/settings.json.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

