Skill Evaluate — Development skill for Claude Code
Evaluate whether a skill amendment improved outcomes (self-improving skills EVALUATE layer).
How to install Skill Evaluate
Installs to ~/.claude/skills/unisone-self-improving-skills-skill-evaluate/SKILL.md
mkdir -p ~/.claude/skills/unisone-self-improving-skills-skill-evaluate && curl -fsSL https://raw.githubusercontent.com/unisone/self-improving-skills/HEAD/commands/skill-evaluate.md -o ~/.claude/skills/unisone-self-improving-skills-skill-evaluate/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Skill Evaluate does
description: Evaluate whether a skill amendment improved outcomes (self-improving skills EVALUATE layer) argument-hint: allowed-tools: Read, Glob, Grep, Bash
Evaluate Skill: $ARGUMENTS (Self-Improving Skills — EVALUATE)
This is the EVALUATE step — the final gate in the Observe→Inspect→Amend→Evaluate loop. A self-improving system must never be trusted simply because it can modify itself. Every amendment must be evaluated.
Step 1: Find the Amendment Record
Search for
Alternatives in Development
- Skill Creator — Create new skills, modify existing skills, and measure skill performance 94.1k ★
- Claude Md Management — Tools for managing and improving CLAUDE.md files 14k ★
- Route Eval — /route-eval — Evaluate a project and set its routing profile 705 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Improve Skill
Propose and apply targeted improvements to a specific Claude Code skill (self-improving skills AMEND layer)
Skill Feedback
Rate a skill's performance after use (self-improving skills OBSERVE layer — user feedback)
Meta Improve
Analyze evaluator feedback and agent run failures, then rewrite command templates to prevent recurring issues.
Vibememo Eval
Periodic VibeMemo evaluation and capture. Runs on a loop (default 30m) to assess whether significant decisions
Grow Team
Invoke Org Designer to specifically evaluate whether the team needs a new role, a split, or activation of a pl
Should I Start Fresh
Evaluate whether the current conversation should continue or whether starting a new session would produce bett
Related Agents
Aidd Escape Analyst
Analyzes a defect that escaped to production — which verification layer should have caught it, why it did not,
Product Data Analyst
Sorcerer — Interpret product usage, conversion, drop-off, errors and feedback; form evidence-backed hypotheses
Agent Improver
Use this agent to analyze agent performance, identify patterns in agent outputs, and generate improvement reco
