Harness Evaluator — Development agent for Claude Code
Use this agent when the harness needs to evaluate a generator's output.
How to install Harness Evaluator
Installs to ~/.claude/agents/vladaimanager-claude-harness-harness-evaluator.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/VladAIManager/claude-harness/HEAD/agents/harness-evaluator.md -o ~/.claude/agents/vladaimanager-claude-harness-harness-evaluator.md Restart Claude Code, or start a new session, for it to be picked up.
What Harness Evaluator does
name: harness-evaluator description: Use this agent when the harness needs to evaluate a generator's output. The evaluator interacts with the running application like a user — clicking, navigating, testing features — and produces a structured eval report with pass/fail verdicts per criterion. Uses Playwright MCP when available.
Context: The generator has completed a build round and the harness needs QA. user: "Build round 1 is complete. Evaluate the output." assistant: "DispatchinAlternatives in Development
- Recon Ranker — Attack surface ranking agent 4.5k ★
- Example Curator — Use this agent when you need to evaluate CLAUDE.md examples for inclusion in the awesome-claude-md repository 570 ★
- Loop Execution Evaluator 334 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Gan Evaluator
GAN Harness — Evaluator agent. Tests the live running application via Playwright, scores against rubric, and p
Harness Generator
Use this agent when the harness orchestrator dispatches implementation work. The generator builds the applicat
Harness Planner
Use this agent when the harness orchestrator needs to expand a brief user prompt into a detailed product speci
Character Generator
Build one whole, grounded character from the world down. Given the world's present systems and a target (the r
Gad Systematic Source Evaluator
Identifies and evaluates systematic uncertainty sources -- both experimental (JES, JER, b-tag, lepton ID, lumi
Attack Ideator
Phase 10 Review Chamber creative attack hypothesis generator that thinks like a hacker, chains low-severity is
Related Skills
Claude Technique Evaluator
Evaluate new Claude prompting patterns, tools, or workflow changes. Produces go/no-go recommendations and inte
GitHub Readme Writer
An AI agent skill that reads every file in a repo before writing a word, then produces a README with real comm
Guide Nt
Generate a searchable single-file HTML guide — walk each role's features in a real browser capturing screensho
