Loading...
Loading...
Proceed with implementation based on the TODO list in Spec.md. Execute review→check→commit after each task is completed.
npx skill4agent add myuon/ccskills impl/spec/specheader: "Review"
question: "How will code reviews be conducted?"
options:
- label: "Self-review (Default)"
description: "Check from the perspectives of DRY, complexity, specification compliance, and security"
- label: "/review スキルを使用"
description: "Invoke the project's review skill"
- label: "Skip"
description: "Proceed without review"header: "Check"
question: "Please provide the pre-commit check command"
options:
- label: "Auto-detection"
description: "Detect from package.json, Makefile, etc."
- label: "npm run lint && npm run test"
description: "Standard npm scripts"
- label: "make check"
description: "check target in Makefile"implement → review → check → commit → next| Aspect | Check Items |
|---|---|
| DRY | Is there any unnecessary code duplication? |
| Complexity | Is the approach overly complex for the intended purpose? |
| Specification Compliance | Does it meet the requirements in Spec.md? |
| Avoid Ad-hoc Solutions | Is it not an ad-hoc fix? |
| Security | Does it avoid anti-patterns such as those in OWASP Top 10? |
/review/commitfeat(auth): Add login screen prototype
WIP: Some tests fail as implementation is scheduled for subsequent tasks/impl/impl| File | Commands to Detect |
|---|---|
| lint, typecheck, build, test in scripts |
| lint, check, build, test targets |
| ruff, mypy, pytest, etc. |
| cargo check, cargo test, cargo clippy |