agent-wiki-compare-out…
Compare successful and failed normalized agent trajectories to derive evidence-backed agent-wiki guidelines. Use when Codex has…
Diagnose evolve health on Claude — verify the CLAUDE.md @import is actually loading the thin EVOLVE.md into sessions
Claude Code
$ npx -y skills add AgentToolkit/altk-evolve --skill doctor --agent claude-code
Codex
$ npx skills add AgentToolkit/altk-evolve --skill doctor -a codex -y
Cursor
$ npx skills add AgentToolkit/altk-evolve --skill doctor -a cursor -y
Gemini CLI
$ npx skills add AgentToolkit/altk-evolve --skill doctor -a gemini-cli -y
GitHub Copilot
$ npx skills add AgentToolkit/altk-evolve --skill doctor -a github-copilot -y
OpenCode
$ npx skills add AgentToolkit/altk-evolve --skill doctor -a opencode -y
Windsurf
$ npx skills add AgentToolkit/altk-evolve --skill doctor -a windsurf -y
name: doctor description: Diagnose evolve health on Claude — verify the CLAUDE.md @import is actually loading the thin EVOLVE.md into sessions context: fork
On Claude, evolve is delivered by a single `@.evolve/EVOLVE.md` import line in this repo's `./CLAUDE.md`. That import requires a one-time, per-project "allow external imports" approval. If you (or a teammate) declined it — even once, in a past session — Claude silently disables the import forever, the thin EVOLVE.md never loads, and evolve becomes a no-op with **no error**.
This skill checks whether the import is actually reaching your sessions. The installed EVOLVE.md carries a canary line, and only you can tell whether it reached your context: Claude does not write CLAUDE.md or its imports into session transcripts, so there is nothing on disk to check instead.
1. Look in the project instructions already loaded into your context (the `CLAUDE.md` contents and anything it imports) for a line that starts with `Evolve import canary:`. Do **not** read, open or search any file for it — reading `.evolve/EVOLVE.md` would find the line whether or not the import works, and defeat the check.
2. Run the doctor script from the repo root, passing what you found. If you found the line, pass the `EVOLVE_IMPORT_CANARY_…` token from it:
python3 ${CLAUDE_PLUGIN_ROOT}/skills/evolve-lite/doctor/scripts/doctor.py --seen <token>If you did not find it:
python3 ${CLAUDE_PLUGIN_ROOT}/skills/evolve-lite/doctor/scripts/doctor.py --not-seenIt is read-only and always exits 0. Read the status code it prints:
not reaching sessions (you likely declined the external-import approval). Follow the remediation the script prints: purge the project approval, start a new session, and **Allow** the import dialog.
canary, or this session loaded an older copy; re-run the installer to refresh it, then start a new session.
steps above.
Relay the status and any remediation to the user.
Evolve helps AI agents learn from experience. It turns past trajectories into reusable guidelines that help agents complete future tasks, without changing model weights.
Compare successful and failed normalized agent trajectories to derive evidence-backed agent-wiki guidelines. Use when Codex has…
Read all atomic guidelines in wiki-twobatch/guidelines/ and propose themed clusters that group near-duplicates. Writes cluster…
Consult an agent-wiki for guidelines relevant to the task at hand. The wiki itself documents how to retrieve from it (AGENTS.md).…
Read a normalized Claude Code trajectory JSON and extract reusable guidelines into wiki-twobatch/guidelines/. Use when mining…
Ingest one or more agent trajectories (raw bob/claude traces or normalized JSON) into an agent-wiki end-to-end — convert,…
Read a normalized Claude Code trajectory JSON and write an episodic summary page to wiki-twobatch/summaries/. Use when…
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or…
Ask which skill or flow fits your situation. A router over the skills in this repo.