SkillCompass – Diagnose and Improve AI Agent Skills Across 6 Dimensions
Six-dimension scoring framework beats guesswork for improving Claude Code skills.
Evaluate agent skill quality. Find the weakest link. Fix it. Prove it worked.
Six-dimension scoring for Claude skills when nothing else measures quality this way.
Claude Code users maintaining custom agent skills
Six-dimension scoring framework beats guesswork for improving Claude Code skills.
Fourteen parallel Claude agents grade your startup idea's evidence before you quit your job.
Context-blind auditors catch assumption gaps, but plugin lock-in and single-model limit audience.
AI agents interrogating other AI agents is a genuinely novel vendor evaluation approach.
Security scanning catches data exfiltration before skills go live.
Claude Skill for agent evals, but LangSmith and Arize already own this.