skill-validator — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited skill-validator (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
A systematic quality-assurance workflow for “skills” (markdown-based instruction packs) so they ship clean, usable, and reproducible.
This validator is intentionally opinionated about structure and completeness because the fastest way to make a skill unusable is to ship it with:
Use this validator:
Do not use this as a replacement for functional testing. This validates structure, clarity, and packaging, not runtime correctness.
A validation report with:
Plus a complete, consistent documentation package ready for GitHub.
--- on line 1--- before any markdown content---name (kebab-case, no spaces)description (clear, under ~300 characters)metadata.creatormetadata.version (semantic versioning)license (if distributing publicly)Only use these root keys:
name, description, license, allowed-tools, compatibility, metadataIf you need custom fields, nest them under metadata.
#, ##, ###)-, *, 1.)A publishable skill package includes these files:
Checks:
For public GitHub distribution, keep the language:
Checks:
skill-name/
├── SKILL.md
├── README.md
├── EXAMPLES.md
├── RESEARCH.md
├── CHANGELOG.md
└── TESTING-NOTES.mdevals/ for automated checksscripts/ for helpersreferences/ for PDFs or supporting materialsMark each item as ✅ / ⚠️ / ❌.
Resolve anything that would:
Warnings aren’t blockers, but they often become bug reports later.
If anything is missing, generate it before release.
Use Keep a Changelog structure. Include dates and clear notes.
Create the release, tag the version, and ensure the repo root is readable for first-time visitors.
# VALIDATION REPORT: <SKILL NAME>
**DATE:** <YYYY-MM-DD>
**SKILL VERSION:** <X.Y.Z>
**VALIDATOR VERSION:** <X.Y.Z>
## YAML FRONTMATTER
| CHECK | STATUS | NOTES |
|------|--------|------|
| DELIMITERS PRESENT | ✅/⚠️/❌ | |
| PARSES CLEANLY | ✅/⚠️/❌ | |
| REQUIRED FIELDS PRESENT | ✅/⚠️/❌ | |
| ROOT KEYS VALID | ✅/⚠️/❌ | |
## MARKDOWN TECHNICAL QUALITY
| CHECK | STATUS | NOTES |
|------|--------|------|
| HEADER HIERARCHY | ✅/⚠️/❌ | |
| CODE FENCES | ✅/⚠️/❌ | |
| LINKS VALID | ✅/⚠️/❌ | |
| TABLES RENDER | ✅/⚠️/❌ | |
## CONTENT QUALITY
| CHECK | STATUS | NOTES |
|------|--------|------|
| TRIGGERS CLEAR | ✅/⚠️/❌ | |
| STEP-BY-STEP INSTRUCTIONS | ✅/⚠️/❌ | |
| EXAMPLES SUFFICIENT | ✅/⚠️/❌ | |
## DOCUMENTATION
| CHECK | STATUS | NOTES |
|------|--------|------|
| ALL 6 FILES PRESENT | ✅/⚠️/❌ | |
| TROUBLESHOOTING INCLUDED | ✅/⚠️/❌ | |
| CHANGELOG UPDATED | ✅/⚠️/❌ | |
## OVERALL RESULT
**RESULT:** PASSED / PASSED WITH WARNINGS / FAILED
**FAILURES TO FIX:**
- <LIST>
**WARNINGS TO CONSIDER:**
- <LIST>
**NOTES:**
- <ANYTHING IMPORTANT>FIX: Ensure the file starts with --- on line 1 and closes YAML with ---.
FIX: Move extra fields under metadata.
FIX: Ensure the anchor matches the real header text.
FIX: Add examples with context + inputs + expected outputs + variations.
FIX: Replace “do X” with explicit steps, including filenames/commands where relevant.
© 2026 STUDIO BRITTANY™ https://studiobrittany.com
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.