8363750c13
- Add skill-reviewer v1.0.0 for reviewing Claude Code skills against best practices - Self-review mode: validate your own skills before publishing - External review mode: evaluate others' skill repositories - Auto-PR mode: fork, improve, submit PRs with additive-only changes - Auto-install dependencies: automatically installs skill-creator if missing - Add github-contributor v1.0.0 for strategic open-source contribution - Four contribution types: Documentation, Code Quality, Bug Fixes, Features - Project selection criteria and red flags - PR excellence workflow and reputation building ladder - GitHub CLI commands and conventional commit format - Update marketplace to v1.22.0 with 30 skills - Update documentation (README, README.zh-CN, CLAUDE.md, CHANGELOG) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2.5 KiB
2.5 KiB
Skill Evaluation Checklist
Complete checklist for evaluating Claude Code skills against best practices.
YAML Frontmatter
namefield present and valid- Max 64 characters
- Lowercase letters, numbers, hyphens only
- No reserved words (anthropic, claude)
descriptionfield present and valid- Non-empty
- Max 1024 characters
- Third-person voice
- Includes trigger conditions ("Use when...")
Description Quality
Third-Person Voice Check
❌ "Browse YouTube videos..."
❌ "You can use this to..."
❌ "I can help you..."
✅ "Browses YouTube videos..."
✅ "This skill processes..."
Trigger Conditions Check
Description should include:
- What the skill does
- When to use it
- Specific triggers (file types, keywords, scenarios)
❌ "Processes PDFs"
✅ "Extracts text and tables from PDF files. Use when working with PDF files or when the user mentions PDFs, forms, or document extraction."
Instruction Quality
- Imperative/infinitive form used (verb-first)
- Concise (avoid obvious explanations)
- Clear workflow steps
- Checklist pattern for complex tasks
Imperative Form Check
❌ "You should run the script..."
❌ "The user can configure..."
✅ "Run the script..."
✅ "Configure by editing..."
Progressive Disclosure
- SKILL.md body under 500 lines
- Detailed content in
references/ - Large files include grep patterns
- No duplication between SKILL.md and references
Bundled Resources
Scripts (scripts/)
- Executable with proper shebang
- Explicit error handling (no bare except)
- Clear documentation
- No hardcoded secrets
References (references/)
- Self-explanatory filenames
- Loaded as needed, not always
- No duplication with SKILL.md
Assets (assets/)
- Used in output, not loaded into context
- Templates, images, boilerplate
Privacy and Paths
- No absolute user paths (
/Users/username/) - No personal/company names
- No hardcoded secrets
- Relative paths only
Workflow Pattern
- Clear sequential steps
- Copy-paste checklist provided
- Validation/verification steps included
Error Handling
- Scripts have specific exception types
- Error messages are helpful
- Recovery paths documented
Summary Table
| Category | Status | Notes |
|---|---|---|
| Frontmatter | ||
| Description | ||
| Instructions | ||
| Progressive Disclosure | ||
| Resources | ||
| Privacy | ||
| Workflow | ||
| Error Handling |