ワンクリックでManusで任意のスキルを実行

commit-double-check

スター26

フォーク7

更新日2026年2月2日 07:14

Review commit(s) for correctness, risk, and completeness. Use when the user asks to double-check a commit, review recent commits, or sanity-check before push/merge.

インストール

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。

Manusで実行

ソース

vana-com

vana-com/data-connect

GitHub リポジトリを開く Creator のリポジトリを見る

ダウンロード

Manusで実行

Commit Double Check

Scope

Read-only review. Do not commit, amend, rebase, or push.
Prefer the most capable model available; avoid fast mode for review.

Default scope

If the user does not specify, review the last 3 commits on the current branch.
If there are uncommitted changes, call out that the review may be stale and ask whether to include those diffs.

Scope parsing

Accept user overrides:

"last commit" or "last 1" -> HEAD~1..HEAD
"last N commits" -> HEAD~N..HEAD
"range A..B" or "A...B" -> use that range verbatim
"commit " -> single commit

If both range and count are provided, prefer the explicit range.

Evidence collection (git)

Use shell commands to collect context (read-only):

git status -sb
git log -n 3 --oneline (or the user-provided count/range)
For each commit: git show --stat <sha> then git show <sha>
For a range: git log --oneline <base>..<head> then git diff <base>..<head>

Methodical review loop

For each commit (in order):

Summarize intent from message + diff.
Apply checklist.
Record findings with file(s), why, fix.

React/TypeScript emphasis

When any .tsx, .ts, .jsx, or .js files are touched:

Load vercel-react-best-practices and apply relevant rules.
If unclear, read vercel-react-best-practices/AGENTS.md or specific vercel-react-best-practices/rules/*.md entries that match the change.

Review checklist

Correctness: logic, edge cases, null handling, invariants
Data safety: migrations, deletes, backwards compatibility
Security: auth, input validation, secrets
Performance: obvious hot paths, expensive loops, large payloads
Reliability: retries, timeouts, error handling
UX/API: breaking changes, schema mismatches, docs
Tests: missing coverage or brittle tests called out
Tooling/config: dependency changes, build/test scripts, env vars
React/TypeScript focus (when applicable):
- Hooks: deps arrays, stale closures, cleanup
- Rendering: unnecessary re-renders, memo usage, key stability
- Types: unsafe any, missing null guards, widening
- Props/state: mutation, derived state pitfalls

Output format

Findings first, ordered by severity (Critical/Major/Minor).
Each finding: short title + file(s) + why it matters + suggested fix.
Then assumptions/questions.
Then a 1-2 line summary + suggested test commands.

Example triggers

"double check this commit"
"review last 3 commits"
"sanity check before push"

このリポジトリの他の Skills

同じリポジトリ

patch-first-debugging

vana-com/data-connect

Debug and repair code using minimal, reviewable patches instead of full-file rewrites. Use this whenever the user is fixing a bug, iterating on failing tests, asks for a surgical code change, wants a minimal diff, or is in a tight debug loop where output size and token cost matter. Prefer this even if the user does not explicitly say "patch."

2026-03-2526

test-fix-loop

vana-com/data-connect

Build or use a tight automated test-fix loop that repairs one failing test at a time with minimal patches and narrow context. Use whenever the user wants a repair harness, constrained bug-fixing loop, failing-test repair workflow, or a prompt that turns the model into a surgical fix engine instead of a general coding assistant.

2026-03-2526

chrome-cdp

vana-com/data-connect

Interact with local Chrome browser session (only on explicit user approval after being asked to inspect, debug, or interact with a page open in Chrome)

2026-03-2426

cognitive-load

vana-com/data-connect

Reduce cognitive load when implementing or refactoring code for readability.

2026-03-2426

conventional-commits

vana-com/data-connect

Use Conventional Commits when creating reviewing or fixing commit messages.

2026-03-2426

ui-component-audit

vana-com/data-connect

Audit and refactor React UI components to match DataConnect UI implementation rules (Text component, tokens, Tailwind, naming) and Vercel React/composition guidance. Use when the user asks for a UI audit, React audit, composition review, or to fix UI implementation issues in components.

2026-03-0926

name	commit-double-check
description	Review commit(s) for correctness, risk, and completeness. Use when the user asks to double-check a commit, review recent commits, or sanity-check before push/merge.

Commit Double Check

Scope

Read-only review. Do not commit, amend, rebase, or push.
Prefer the most capable model available; avoid fast mode for review.

Default scope

If the user does not specify, review the last 3 commits on the current branch.
If there are uncommitted changes, call out that the review may be stale and ask whether to include those diffs.

Scope parsing

Accept user overrides:

"last commit" or "last 1" -> HEAD~1..HEAD
"last N commits" -> HEAD~N..HEAD
"range A..B" or "A...B" -> use that range verbatim
"commit " -> single commit

If both range and count are provided, prefer the explicit range.

Evidence collection (git)

Use shell commands to collect context (read-only):

git status -sb
git log -n 3 --oneline (or the user-provided count/range)
For each commit: git show --stat <sha> then git show <sha>
For a range: git log --oneline <base>..<head> then git diff <base>..<head>

Methodical review loop

For each commit (in order):

Summarize intent from message + diff.
Apply checklist.
Record findings with file(s), why, fix.

React/TypeScript emphasis

When any .tsx, .ts, .jsx, or .js files are touched:

Load vercel-react-best-practices and apply relevant rules.
If unclear, read vercel-react-best-practices/AGENTS.md or specific vercel-react-best-practices/rules/*.md entries that match the change.

Review checklist

Correctness: logic, edge cases, null handling, invariants
Data safety: migrations, deletes, backwards compatibility
Security: auth, input validation, secrets
Performance: obvious hot paths, expensive loops, large payloads
Reliability: retries, timeouts, error handling
UX/API: breaking changes, schema mismatches, docs
Tests: missing coverage or brittle tests called out
Tooling/config: dependency changes, build/test scripts, env vars
React/TypeScript focus (when applicable):
- Hooks: deps arrays, stale closures, cleanup
- Rendering: unnecessary re-renders, memo usage, key stability
- Types: unsafe any, missing null guards, widening
- Props/state: mutation, derived state pitfalls

Output format

Findings first, ordered by severity (Critical/Major/Minor).
Each finding: short title + file(s) + why it matters + suggested fix.
Then assumptions/questions.
Then a 1-2 line summary + suggested test commands.

Example triggers

"double check this commit"
"review last 3 commits"
"sanity check before push"