Harness Checker — Development agent for Claude Code
Fresh-context artifact evaluator for goal loops.
How to install Harness Checker
Installs to ~/.claude/agents/leadgrowgtm-loop-engineer-harness-checker.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/LeadGrowGTM/loop-engineer/HEAD/.claude/agents/harness-checker.md -o ~/.claude/agents/leadgrowgtm-loop-engineer-harness-checker.md Restart Claude Code, or start a new session, for it to be picked up.
What Harness Checker does
name: harness-checker description: Fresh-context artifact evaluator for goal loops. Reads only final artifacts — never Maker reasoning or PROGRESS.md. Scores each rubric dimension with file:line evidence citations. Writes CYCLE_LOG.md. Decides PASS, ITERATE, or PLATEAU. Use in the Checker phase of any harness eval loop. tools: Read, Glob, Write model: claude-sonnet-5
You are the Harness Checker. You are at depth level 4 (goal=0, planner=1, maker=2, prover=3, checker=4).
**You did NOT w
Alternatives in Development
- Agent Refactorer — Principal Clean Code Specialist specializing in code simplification, performance tuning, and refactoring loops 957 ★
- Goal Monitor — PROACTIVELY use this agent for executive function providing direction and accountability 52 ★
- Eval Judge — Use this agent during the /eval Skill Phase 3 (Epic #803, issue #810) to judge — from a session-eval record's 49 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Harness Maker
Executes a goal plan phase by phase using skills from PLAN.md routing. Commits at each phase boundary. Writes
Harness Prover
Running-app verifier for harness eval loops. Drives the live feature (browser, API, or CLI) and returns a bina
Gan Evaluator
GAN Harness — Evaluator agent. Tests the live running application via Playwright, scores against rubric, and p
Resume Judge
Scores a resume or letter draft against the anchored rubric in templates/rubric-resume.md and returns per-dime
Oracle Evaluator
Runs the evaluation harness on candidate artifacts via pp_harness best-of-N and Borda. Invoke after a draft ha
Output Auditor
The quality gate. Scores artifacts against a published rubric, verifies every citation against the ledger, and
Related Skills
Rubric Evaluator
Evaluate Claude Code and Codex skill directories with deterministic rubric checks and graded, fixable reports.
Evaluator Library
Dispatch a fresh-context evaluator against a domain rubric. Bundled domains: code-quality, ui-design, prose, s
Skill Checker
🔍 Claude Code skill evaluator — check any skill's functionality, security, dependencies, API needs, and token