64 lines
2.1 KiB
Markdown
64 lines
2.1 KiB
Markdown
# TOOLS.md - Local Notes
|
|
|
|
Skills define _how_ tools work. This file is for _your_ specifics — the stuff that's unique to your setup as a code reviewer.
|
|
|
|
## What Goes Here
|
|
|
|
- Linter configurations and rulesets
|
|
- Static analysis tools and their settings
|
|
- Test coverage thresholds and quality gates
|
|
- Code style guides ( naming conventions, formatting standards)
|
|
- Security scanning tools and vulnerability databases
|
|
- Performance benchmarking tools
|
|
- Architecture decision records (ADR) references
|
|
|
|
## Examples
|
|
|
|
```markdown
|
|
### Linters
|
|
|
|
- ESLint: `extends: ['eslint:recommended', '@company']`
|
|
- RuboCop: `target_ruby_version: 3.2`
|
|
- Pylint: `max-line-length = 100`
|
|
|
|
### Testing Standards
|
|
|
|
- Minimum coverage: 80% for new code, 70% overall
|
|
- Required: unit tests + integration tests for API changes
|
|
- Test framework: Jest for JavaScript, pytest for Python
|
|
|
|
### Architecture
|
|
|
|
- Follow layered architecture: presentation → business → data
|
|
- Use dependency injection; no service locator patterns
|
|
- Feature flags for all experimental code
|
|
```
|
|
|
|
## Why Separate?
|
|
|
|
Skills are shared. Your setup is yours. Keeping them apart means you can update your review criteria without affecting other agents, and maintain your specific code quality standards.
|
|
|
|
## Code Review Focus Areas
|
|
|
|
When reviewing, always check:
|
|
|
|
1. **Plan alignment**: Does the implementation match the original design?
|
|
2. **Security**: Input validation, injection risks, auth/authorization
|
|
3. **Performance**: Algorithm complexity, unnecessary computations, memory leaks
|
|
4. **Maintainability**: Readability, naming, duplication, coupling
|
|
5. **Testing**: Coverage, edge cases, test quality
|
|
6. **Documentation**: Comments, README updates, API docs
|
|
|
|
## ClawTeam Integration
|
|
|
|
When coordinating with other agents via ClawTeam:
|
|
|
|
- Use `clawteam task create` to assign review tasks if splitting work
|
|
- Use `clawteam inbox send` to ask clarifying questions to the developer
|
|
- Use `clawteam board show` to track multiple reviews in progress
|
|
- Use `clawteam workspace merge` to integrate review feedback into worktrees
|
|
|
|
---
|
|
|
|
Add whatever helps you maintain high code quality. This is your checklist.
|