메뉴
Turn an ambiguous request into a grounded plan, an evidence-backed execution, and a reviewed result. Use for non-trivial coding tasks that benefit from a plan→execute→review loop.
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/w7-mgfcode/intentguard-docs-repo --skill plan-execute-review명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
Manage GitHub repositories end to end - issues, labels, pull request workflows, branch protection, semantic-version releases, GitHub Actions CI, and deployment to Vercel, Google Cloud Run, and AWS. Use when setting up or fixing CI workflows, writing issue or PR templates, cutting a release, configuring branch protection or CODEOWNERS, pinning or upgrading GitHub Actions, wiring keyless OIDC authentication from Actions to a cloud provider, or building a deploy pipeline. Do NOT use for decomposing work into milestone, umbrella, epic, or subtask hierarchies - see github-work-decomposition. Do NOT use for AWS-native CodePipeline, CodeBuild, or CodeDeploy orchestration - see the aws-deployment skill.
Decomposes large GitHub work into a milestone → umbrella → epic → subtask hierarchy using gh CLI, native sub-issues, issue types, and Projects v2, with a deterministic slice-scoring engine that decides how many children each level should have. Use when breaking down a roadmap item, planning a milestone, splitting an umbrella into epics, splitting an epic into subtasks, right-sizing a backlog, or deciding how many pieces work should be cut into. Triggers include: 'create a milestone with umbrellas', 'decompose this umbrella into epics', 'split this epic into subtasks', 'how many slices should this be', 'break down this feature', 'right-size this backlog'. Do NOT use for authoring issue prose alone, triaging existing issues, release automation, or CI workflow authoring.
Generate and maintain a repository's agent-instruction files — AGENTS.md (canonical), CLAUDE.md and GEMINI.md (thin shims that import AGENTS.md), README.md (for humans), and an optional llms.txt — with cross-file consistency, single-source-of-truth (no duplicated policy), and a validation gate. Use when the user asks to set up, create, generate, audit, fix, align, or update AGENTS.md / CLAUDE.md / GEMINI.md / README.md / llms.txt, make a repo "agent-ready", or resolve duplication/drift between these files. Reconciles existing files via diff-then-approve; never clobbers human-written content.
SOC 직업 분류 기준
SKILL.md 표시 중
| name | plan-execute-review |
| description | Turn an ambiguous request into a grounded plan, an evidence-backed execution, and a reviewed result. Use for non-trivial coding tasks that benefit from a plan→execute→review loop. |
A portable, provider-agnostic workflow for taking an ambiguous request to a reviewed change.
This is a seed skill demonstrating the .agents/skills/ convention; extend it for your needs.
Read assets/prompts/plan.md. Ground the request in the actual repo: list the files involved,
state assumptions explicitly, and produce a numbered, verifiable plan. Stop for approval if the
request is underspecified.
Follow assets/prompts/execute.md. Make the smallest change that satisfies the plan. Prefer
worktree isolation for anything that mutates files (see docs/safe-execution.md). Run the
repo's verify command after each meaningful step.
Use assets/prompts/review.md. Check the change against the plan and the repo's definition of
"done" (the verify command in AGENTS.md). Surface anything skipped or uncertain.
The verify command passes and the change matches the approved plan.
