92
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill jmagly/aiwg --skill quality-checker- SKILL.md8.0 KB
Overview
This skill validates a skill package for quality, completeness, and standards compliance before packaging. It scores structure, content, code examples, and documentation, then emits a pass/warn/fail decision and actionable recommendations. Use it as a pre-release gate to catch gaps and enforce minimum quality thresholds.
How this skill works
The checker inspects the skill directory layout, required documentation files, reference material, and embedded code examples. It measures content metrics (length, sections, links), verifies example counts and language tags, and evaluates reference file coverage. Results are aggregated into a weighted score and a machine-readable quality report saved to a checkpoints folder. Validation can run at quick, standard, full, or strict levels and supports custom rules and thresholds.
When to use it
- Before packaging or publishing a skill to ensure it meets release standards
- As an automated CI step to gate merges that affect skill artifacts
- When onboarding contributors to verify documentation and example completeness
- To audit third-party or legacy skills for gaps and modernization needs
Best practices
- Define clear quality criteria and validation level (quick/standard/full/strict) before running checks
- Require a concise manifest and navigation section so automated parsing finds key sections
- Include at least three real, language-tagged code examples with minimal placeholders
- Maintain multiple reference files with substantive content rather than single sparse documents
- Treat WARN results as review items and FAIL results as blockers until corrected
Example use cases
- Automated pre-release CI job that fails the pipeline when the quality score is below threshold
- Local developer command to run a quick validation before opening a pull request
- Quality audit of a skill catalog to produce prioritized remediation tickets
- Custom validation enforcing API coverage rules for reference documentation
FAQ
PASS means the score meets the configured pass threshold and the package is ready for packaging; WARN indicates reviewable issues; FAIL requires remediation before packaging.
Can I customize validation rules and thresholds?
Yes. You can configure thresholds, minimum lines, minimum example counts, required reference files, and add custom grep-style rules for specialized checks.