Evol-aiEvol-aiOfficialยท1 Agent Skills Included

SkillCompass

Skill quality scoring, security audits, and guided improvement

Evaluates skill quality across six weighted dimensions including structure, triggers, security, and functionality. Pinpoints the weakest dimension, applies targeted fixes, and re-verifies results so no change is saved unless it actually helps. Tracks usage to surface idle, stale, or risky skills, with snapshots and rollback for safe version management.
npx skills add Evol-ai/SkillCompass --all -g -y

All Skills in This Repository (1)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install SkillCompass?โ–ผ

Run `npx skills add Evol-ai/SkillCompass --all -g -y` in your terminal to install it globally for your agent.

What does SkillCompass evaluate?โ–ผ

It scores skills on six dimensions: structure, trigger accuracy, security, functional quality, comparative value, and uniqueness, then gives a PASS, CAUTION, or FAIL verdict.

Can SkillCompass fix weak skills automatically?โ–ผ

Yes. The eval-improve command targets the weakest dimension, applies a fix, and re-evaluates, only saving the change if the score actually improved without regressions.

Does SkillCompass work with Claude Code and OpenClaw?โ–ผ

Yes. It runs locally in Claude Code and OpenClaw, and also supports 45+ agents including Cursor, Codex, and Gemini CLI.

Is my skill data sent anywhere?โ–ผ

No. SkillCompass is local-first: all evaluations, snapshots, and usage tracking stay on your machine, with network access only when you explicitly request updates.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’