メニュー
Bidirectional AI slop scanner — Claude + GPT independently analyze, then debate disagreements.
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
| name | cleanup |
| description | Bidirectional AI slop scanner — Claude + GPT independently analyze, then debate disagreements. |
| user-invocable | true |
/cleanup [scope] where scope is: deps, unused-exports, hardcoded, duplicates, deadcode, or all
Claude analyzes independently, then codemoot cleanup runs deterministic regex + GPT scans. 3-way merge with majority-vote confidence.
Scan the codebase yourself using Grep/Glob/Read. For each scope:
Save findings as JSON to a temp file.
codemoot cleanup --scope SCOPE --host-findings /path/to/claude-findings.json
Show summary: total, high confidence, disputed, adjudicated, by source.
For Claude/GPT disagreements, optionally debate via codemoot debate turn.
Autonomous build loop — debate, plan, implement, review, fix — all in one session with GPT review.
Get an independent GPT review via Codex CLI. Use when you want a second opinion on code, plans, or architecture from a different AI model.
Real Claude vs GPT multi-round debate. Use when you need a second opinion, want to debate architecture decisions, or evaluate competing approaches with multi-model collaboration.
Send execution plans to GPT for structured review via Codex CLI. Use when you have a written plan and want independent validation before implementation.