Test Quality Adversary — Testing & QA agent for Claude Code
Fresh-context, refute-framed review of what the TESTS actually promise — the head-of-QA loss function.
How to install Test Quality Adversary
Installs to ~/.claude/agents/davalst-tdd-playbook-test-quality-adversary.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/davalst/tdd-playbook/HEAD/.claude/agents/test-quality-adversary.md -o ~/.claude/agents/davalst-tdd-playbook-test-quality-adversary.md Restart Claude Code, or start a new session, for it to be picked up.
What Test Quality Adversary does
name: test-quality-adversary description: Fresh-context, refute-framed review of what the TESTS actually promise — the head-of-QA loss function. Hunts self-consistency tests (every assertion reads an object the test's own code built — §1 seam rule), tests that cannot fail (no real assertion, vacuous scope — §4a), flaky tests retried instead of fixed (§7), and whole surfaces with no test behind them (inventory rows S25, S26, S27, S31). Distinct from mutation-runner — a mutation score is blind
Alternatives in Testing & QA
- Consistency QA 1.5k ★
- Review Test Quality — Reviews test code for behavior-driven quality and completeness: whether tests verify real behavior vs chasing 432 ★
- Test Refactor Specialist.Agent — Improves test code quality and maintainability 225 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Memory Hunter
Memory retrieval lane — given a spec (a consumer's question + task), hunts the wiki for the relevant slices an
Spec Fidelity Auditor
Oracle adversary - finds a green, mutation-killing test whose ASSERTION contradicts an authoritative external
Regression Hunter
Use when a test, behavior, or build worked at a known-good commit and is broken at HEAD. Bisects the commit hi
Render QA
Renders a built artboard, records deterministic visual/GIF evidence, and applies render, contrast, seam, motio
Test Hunter
Clawker-specific adversarial subagent that hunts down wasteful, self-serving, and redundant tests in this Go C
Plan Review
Fresh-eyes review of a written PLAN before any code is built. Grounds the plan against the existing codebase a
Related Skills
Audit Mirage Analyze
Phases 2-3 and 7 of auditing-green-mirage: systematic line-by-line audit, the Green Mirage Patterns, named ass
Agent Eval
Head-to-head comparison of coding agents / models on reproducible tasks, with pass-rate / cost / time / consis
Codex Council
Multi-perspective parallel Codex review for Claude Code — N sub-agents, each framed with a role you decide per