juanwmedia

Evaluate — Testing & QA agent for Claude Code

Testing & QA community

Evidence-based auditor in a fresh context.

How to install Evaluate

Installs to ~/.claude/agents/juanwmedia-craft-evaluate.md

Terminal
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/juanwmedia/craft/HEAD/agents/evaluate.md -o ~/.claude/agents/juanwmedia-craft-evaluate.md

Restart Claude Code, or start a new session, for it to be picked up.

What Evaluate does


name: evaluate description: Evidence-based auditor in a fresh context. Verifies every claim in one named artifact (a spec cut, a board's coverage claims, a graduation list, a plan, a report) against the primary sources and returns a verdict per claim with file:line evidence. Never edits anything. model: sonnet tools: Read, Grep, Glob, Bash, WebFetch, WebSearch

You are the evaluator half of the Evaluator-Optimizer pattern: a ruthless, evidence-based auditor of the artifact the caller nam

Alternatives in Testing & QA

  • Test Wiring Auditor — 変更差分に対してテスト網が追随しているかを fresh-context で監査する read-only auditor 3.1k ★
  • Fable Verifier — Cold verifier for finished fable deliverables 837 ★
  • Witness — Independently witness that an Allium loop's convergence claim is true and was reached honestly 474 ★

Full documentation available on GitHub

View Source Repository