スキル一覧に戻る
achhina

finalize

by achhina

A list of my personal configs

0🍴 0📅 2026年1月25日
GitHubで見るManusで実行

SKILL.md


name: finalize description: This skill should be used when the user asks to "finalize", "/finalize", "ship it", "ship this", "wrap up", "wrap this up", or "finish and commit". Orchestrates the complete workflow for finishing development work including running tests, code review, comment cleanup, and creating a commit. version: 0.1.0

Finalize Development Work

This skill orchestrates the complete workflow for finishing development work by running tests, performing code review, cleaning up comments, and creating a well-thought-out commit.

Purpose

When development work is complete, finishing properly requires several quality gates: verifying tests pass, reviewing code quality, removing spurious comments, and creating a meaningful commit. This skill automates this multi-step workflow to ensure consistent quality before committing changes.

When to Use

Use this skill when the user indicates work is ready to be finalized and committed:

  • "finalize" or "/finalize"
  • "ship it" or "ship this"
  • "wrap up" or "wrap this up"
  • "finish and commit"

Workflow Steps

Execute the following steps in order, stopping only if a critical failure occurs that cannot be automatically resolved.

Step 1: Run Tests

Detect and run the appropriate test command for the project:

Detection priority:

  1. Check for package.json with test script → use npm test
  2. Check for pytest.ini, pyproject.toml, or .py test files → use pytest
  3. Check for go.mod → use go test ./...
  4. Check for Cargo.toml → use cargo test
  5. Check for Makefile with test target → use make test

Execution:

# Run the detected test command
<detected-test-command>

On test failure:

  1. Do not proceed with remaining steps
  2. Invoke the superpowers:systematic-debugging skill to analyze and fix the failures
  3. After fixes are applied, re-run tests to verify they pass
  4. If tests still fail after debugging attempt, report to user and halt
  5. Only proceed to Step 2 when all tests pass

Step 2: Code Review

Invoke the code-review agent to analyze recent changes:

Determine review scope:

# Get list of modified files for review
git diff --name-only

Run code review: Invoke the /code-review skill (uses tdd-workflows:code-reviewer agent for comprehensive analysis):

Use Skill tool: skill="code-review"

This provides comprehensive review covering:

  • Security vulnerabilities (OWASP Top 10, injection attacks, secrets)
  • Performance issues (algorithms, database queries, memory management)
  • Code quality (DRY, SOLID, maintainability, complexity)
  • Error handling and resilience
  • Testing and testability
  • Style and conventions

On review findings:

  • Analyze all code review comments and issues identified
  • Apply fixes for all issues found (style violations, potential bugs, best practice deviations)
  • Re-run tests after applying fixes to ensure nothing breaks
  • If any fixes cannot be automatically applied, report to user and halt
  • Only proceed to Step 3 when all review issues are resolved and tests still pass

Step 3: Remove Spurious Comments

Invoke the comments skill to clean up obvious and redundant comments:

Use Skill tool: skill="comments"

The /comments skill launches the comment-remover agent which will:

  • Scan uncommitted changes for obvious comments
  • Remove redundant and self-evident comments
  • Preserve valuable documentation (TODO/FIXME, "why" explanations, docstrings)
  • Runs in isolated context to reduce main conversation context usage

Step 4: Create Commit

Invoke the commit skill to create a well-thought-out commit message:

Use Skill tool: skill="commit"

The /commit skill will:

  • Analyze all changes (git diff and git status)
  • Review recent commit messages for style consistency
  • Draft a meaningful commit message focused on the "why"
  • Create the commit with staged changes

Error Handling

Test failures:

  • Attempt automatic fix using systematic-debugging skill
  • Verify fix with test re-run
  • Halt if tests still fail after debugging

Code review issues:

  • Analyze all findings from the comprehensive code review
  • Apply fixes for all identified issues (security, performance, style, logic)
  • Re-run full test suite after applying fixes
  • Halt if critical issues cannot be auto-fixed
  • Ensure all high-priority review feedback is addressed before proceeding

Comment cleanup:

  • Non-critical, proceed even if no comments removed

Commit creation:

  • If no changes to commit, report and complete successfully
  • If pre-commit hooks fail, analyze and fix issues, then retry commit

Success Criteria

The workflow completes successfully when:

  1. All tests pass
  2. Code review finds no critical issues (or issues are fixed)
  3. Spurious comments are removed
  4. A commit is created with a meaningful message

Notes

  • This workflow assumes work is complete and ready to commit
  • Each step builds on the previous step's success
  • Test failures trigger debugging, not just reporting
  • Code quality gates prevent committing problematic code
  • The final commit represents polished, reviewed, tested work

スコア

総合スコア

50/100

リポジトリの品質指標に基づく評価

SKILL.md

SKILL.mdファイルが含まれている

+20
LICENSE

ライセンスが設定されている

0/10
説明文

100文字以上の説明がある

0/10
人気

GitHub Stars 100以上

0/15
最近の活動

3ヶ月以内に更新がある

0/10
フォーク

10回以上フォークされている

0/5
Issue管理

オープンIssueが50未満

+5
言語

プログラミング言語が設定されている

+5
タグ

1つ以上のタグが設定されている

0/5

レビュー

💬

レビュー機能は近日公開予定です