Red Verifier — Testing & QA agent for Claude Code
在目前 worktree 驗證 target test 為 RED,並產出 policy-resolved evidence.
How to install Red Verifier
Installs to ~/.claude/agents/trionnemesis-ai-nativestdd-vdd-red-verifier.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/trionnemesis/AI-NativeSTDD-VDD/HEAD/.claude/agents/red-verifier.md -o ~/.claude/agents/trionnemesis-ai-nativestdd-vdd-red-verifier.md Restart Claude Code, or start a new session, for it to be picked up.
What Red Verifier does
name: red-verifier description: 在目前 worktree 驗證 target test 為 RED,並產出 policy-resolved evidence。 model: haiku tools:
- Read
- Bash disallowedTools:
- Write
- Edit
主 agent 必須提供 `requirement_id`、exact test path/name、implementation path,以及依 `.vdd/path-policy.json` 解析的 evidence path。
- 讀取
.vdd/path-policy.json;若不存在,使用 compatibility defaults。 - 確認 test path 在 configured
test_roots中,並可被 pytest collect。 - 執行 exact target test,確認目前為真實的需求失敗,不是 ConnectionError/PermissionError。
Alternatives in Testing & QA
- Gitnexus Test CI Verifier — GitNexus test and CI reviewer 45.8k ★
- TDD Orchestrator — TDD orchestrator enforcing red-green-refactor discipline and test-driven development 2.8k ★
- Logic Hacker — Red Team persona for Business Logic and Auth manipulation 548 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Red Quality Gate
Automated quality gate that runs after a failing test is written (RED) and before any implementation (GREEN).
Test Review Gates
Runs the objective test-verification gates (red-green implementation-removal + mutation testing) for ONE shard
Factory Worker
Agent Factory teammate — owns ONE Linear unit at a time end-to-end in its own worktree with strict test-first
Implementation Orchestrator
Use when the root orchestrator needs ONE phase of an arbitrado plan actually built — sequences test-writer (RE
Ac Verifier
Runs the Tech Stack Build, Test, Lint (and optionally Dev) commands listed in a ROADMAP.md against the current
Red First Verifier
Independently PROVE a regression/feature test actually fails without the fix. Use after writing a test-for-a-b
Related Skills
Autopilot Nt
The executor — work a fix-workplan or prose goal to completion unattended in an isolated worktree; verify fres
Harden Config
Deployment hardening step 5 — validate runtime/deploy config per target. Railway: railway.json/toml, healthche
Blade Agent Runtime
Blade Agent Runtime (BAR) - A safe execution runtime for AI coding agents with Git worktree isolation, step le