메뉴
List and reconcile canonical skills across a bounded collection of local repositories. Use when the user asks which skills repositories have, where they occur, or whether a shared skill has conflicting language.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Code review a pull request or the current working changes with independent reviewer lenses, validated findings, and severity-rated results. Establishes a branch and a draft pull request when none exists yet. Accepts an effort level, an optional fix mode that applies the smallest complete correction for each confirmed finding, and an optional comment mode that posts inline review comments, each written with or without a leading double dash. Asks for whichever of effort or modes the invocation did not state. Reports in chat and never posts or fixes unless the matching mode is requested. Use when asked to review a PR, review current changes, or run /code-review.
Coordinate one authorized task across a bounded collection of local repositories while preserving unrelated work. Use when a task spans repositories to apply a scoped change, run the same review or validation, or collect and merge information into one result.
Propagate explicitly requested skills, rules, and their changes across the user's other local repositories, creating missing applicable copies while preserving repository-specific guidance. Use when the user invokes /propagate-skill or $propagate-skill alongside a skill or rule creation or update request.
Upsert `.agents` source-of-truth files for agents, skills, or rules based on user input.
Edit drafts into sharper, more human writing while preserving the writer's personal voice, or detect AI-slop patterns without rewriting. Use when the user wants a draft clearer, more direct, more opinionated, or less AI-sounding, or asks whether writing reads as AI.
Find and watch a GitHub pull request for review findings, investigate each finding, fix and push legitimate issues, and stop once checks are green and review threads are resolved. Use when asked to watch, monitor, poll, babysit, or keep checking a PR for review comments or automated review feedback.
| name | list-skills |
| description | List and reconcile canonical skills across a bounded collection of local repositories. Use when the user asks which skills repositories have, where they occur, or whether a shared skill has conflicting language. |
coordinate-repositories — identifies the bounded repository collection and safely gathers the canonical artifacts.Read each applicable repository's canonical .agents/skills/<name>/SKILL.md files. Exclude repositories whose primary purpose is developing, packaging, testing, or operating the agent-synchronization system itself unless the user explicitly includes them.
For each skill name, merge the repository names that contain it. Compare complete canonical SKILL.md text, including frontmatter. Group identical texts and use the largest group as the baseline. If any other group exists, report every repository outside that baseline as having conflicting language. If two or more groups tie for largest, report all tied groups as conflicting rather than choosing an arbitrary baseline.
Return only these headings and sorted Markdown lists, with no status summary or extra prose:
Skills
- skill-name — (repo-one, repo-two)
## Conflicting Skill Language
- skill-name
- Baseline: repo-one, repo-two
- Different language: repo-three
Omit ## Conflicting Skill Language when no differences exist. If no skills exist, return - None under Skills.