メニュー
Review session context and persist actionable lessons into per-skill evolution.md files. Scriptless — Claude uses native Read/Edit/Write tools directly.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Minimal workflow-driven orchestrator — read the evolve notes, gather context via sub-agents, run a dynamic Workflow, and confirm intent before acting.
Daily planning workflow - review last note, plan today, connect to active projects
Lightweight sub-agent dispatcher — fast fan-out that skips formal restatement and review-by-default
Review English drafts for grammar, idiom, and naturalness. Correct inline with terse explanations.
Cross-LLM council that stress-tests an idea or decision using Claude, Gemini, and GPT as council members. User relays prompts to Gemini and GPT. Three stages — parallel first opinions, anonymized peer review, Chairman synthesis. Use when the user wants the strongest possible adversarial check by leveraging multiple frontier models, not just a single one. Inspired by karpathy/llm-council.
Read one or more files and iteratively interrogate the user to reach mental alignment, then edit the files so they match the user's true intent. Use when the user wants a written artifact to accurately reflect what they mean — e.g. "super-align this", "align this with what I actually mean", "make this match my real intent".
| name | evolve-skills |
| description | Review session context and persist actionable lessons into per-skill evolution.md files. Scriptless — Claude uses native Read/Edit/Write tools directly. |
Read evolution.md in this skill's folder. Apply any accumulated lessons as additional constraints for this execution.
Distills session experience — successful approaches, failures, user corrections, and preferences — into persistent, per-skill evolution.md files so lessons survive across conversations and skill rewrites.
/evolve
Scan the current conversation context for:
Present findings to the user via AskUserQuestion:
Do not persist anything without explicit user confirmation.
For each confirmed skill:
Read existing <skill-folder>/evolution.md (if any)
Deduplicate — skip lessons that already exist in the file
Append new lessons under a dated entry (format below)
Ensure consumption — if the target skill's SKILL.md does not already contain a Phase 0 evolution read instruction, inject this block right after the frontmatter (before the skill's title heading):
# Phase 0 — EVOLVE
Read `evolution.md` in this skill's folder. Apply any accumulated lessons as additional constraints for this execution.
Write using native Write/Edit tools (no scripts, no JSON intermediaries)
# Evolution Log
## 2026-03-04
### Lessons
- [lesson text]
### User Preferences
- [preference text]
### Fixes
- [fix text]
Only include subsections (Lessons / User Preferences / Fixes) that have content. Omit empty subsections.
AskUserQuestion tool to gather explicit direction.