Skill audit report
Measure output quality, don't vibe it: score a generative task with a two-layer grader — deterministic code metrics + per-dimension LLM-as-judge — over a fixed task set, as signed deltas vs a pinned baseline. Grades cost alongside correctness (pass-slow).
OpenAgentSkill Trust Score
The Trust Score helps an agent decide whether a skill is safe enough to shortlist before installation.
GitHub adoption
FAIL30
22 GitHub stars
Stars/forks activity
FAIL32
22 stars, 4 forks; issue activity unavailable in current metadata
Recent maintenance
PASS100
22d since push
License clarity
PASS86
MIT
README/SKILL.md completeness
INFO76
Public metadata needs stronger README/SKILL.md context
Dependency/runtime risk
INFO72
credential or environment access
Install availability
PASS92
npx skills add byerlikaya/claude-starter-kit --skill eval-grader
Install command safety
INFO68
dynamic command execution, standard package or runtime install path
Permission surface
WARN48
secrets or environment access, filesystem or document access
Repository evidence
PASS86
https://github.com/byerlikaya/claude-starter-kit/tree/main/claude-starter/skills/eval-grader
Review status
WARN46
AI review approval is missing
Agent Proven outcomes
INFO54
No agent outcome data yet
Checks
Install path
92
npx skills add byerlikaya/claude-starter-kit --skill eval-grader
Repository
88
https://github.com/byerlikaya/claude-starter-kit/tree/main/claude-starter/skills/eval-grader
License
86
MIT
Maintenance
100
22d since push
AI review
55
Review approval is missing
README/SKILL.md completeness
84
Usable description available
Dependency risk
72
credential or environment access
Install command safety
68
dynamic command execution, standard package or runtime install path
Permission surface
48
secrets or environment access, filesystem or document access
Stars/forks activity
32
22 stars, 4 forks; issue activity unavailable in current metadata
Adoption
42
22 GitHub stars
Financial decision safety
58
Research-only use: do not treat output as financial advice or execute a position without human approval.
Warnings
Method
This report combines public metadata, AI review output, repository freshness, install readiness, OpenAgentSkill events, quality scoring, trust checks, and the agent safety gate. It is not a full source-code security review.
Compare nearby options
Review a branch or diff against repository standards and the originating spec in two independent analysis passes.
169K Stars · Audit report
Platform to build admin panels, internal tools, and dashboards. Integrates with 25+ databases and any API.
41K Stars · Audit report
Implement work from an approved spec or ticket set, run focused and full tests, invoke code review, and commit the result to the current branch.
176K Stars · Audit report