Debate Judge — Testing & QA agent for Claude Code
L0.97 Adversarial Verification — reads all 3 rounds of prosecution and defense transcripts plus the original diff + spec, then emits exactly one of four VERDICT values (PASS REVISE BLOCK DEADLOC.
How to install Debate Judge
Installs to ~/.claude/agents/contact715-jidoka-debate-judge.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/contact715/jidoka/HEAD/.claude/agents/debate-judge.md -o ~/.claude/agents/contact715-jidoka-debate-judge.md Restart Claude Code, or start a new session, for it to be picked up.
What Debate Judge does
name: debate-judge description: L0.97 Adversarial Verification — reads all 3 rounds of prosecution and defense transcripts plus the original diff + spec, then emits exactly one of four VERDICT values (PASS | REVISE | BLOCK | DEADLOCK) with one paragraph of reasoning. Never modifies files. tools: Read, Grep, Bash
Debate Judge
You are the Debate Judge for **this agentic framework**.
Role
L0.97 — Adversarial Verification layer. Final decision-maker in the 3-round debate engine.
Yo
Alternatives in Testing & QA
- Component Config Reader — Extracts the test surface (properties, styles, events, defaults, nested variants) from a ToolJet app-builder c 41k ★
- Arbiter — You are a specialized validation agent 3.6k ★
- Spec Compliance Judge — Use this agent when you need to determine whether a solution correctly implements a specific part of a specifi 173 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Debate Defender
L0.97 Adversarial Verification — reads the prosecution's arguments plus the original diff + spec, constructs t
Design Designer
Creative UI/UX designer on a design sprint team. Produces design proposals across three rounds of increasing c
Tester Lead
Explicit-invocation only (Team Tester, Lead) — dispatched by the Manager to design meaningful test scenarios/e
Test Coverage Mapper
Reducer subagent. Reads every per-node sidecar's tests_coverage_semantic block, joins against the actual test
Ddw Implementer
Implements exactly ONE block of an approved spec, with its tests, in an isolated context. Returns a report of
Ren Planner
Wiki-aware plan decomposer for the execution doctrine's decompose gate. Spawn AFTER a plan or spec is approved
Related Skills
Design Loop
Run the full design iteration loop (harness component 06): planner → generator → four machine gates → critic,
Evolve Skill
SPIKE — GEPA/DSPy-style offline A/B evolution of ONE prompt-only skill's SKILL.md body against a small fixture
Deliberative Refinement
Run your content through multiple rounds where the AI critiques itself from different expert angles (structure