Skill Verifier — Testing & QA agent for Claude Code
Fresh-session surrogate verifier for improve skill rewrites.
How to install Skill Verifier
Installs to ~/.claude/agents/fredrick84823-fstack-skill-verifier.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/fredrick84823/fstack/HEAD/agents/skill-verifier.md -o ~/.claude/agents/fredrick84823-fstack-skill-verifier.md Restart Claude Code, or start a new session, for it to be picked up.
What Skill Verifier does
name: skill-verifier description: Fresh-session surrogate verifier for improve skill rewrites. Use when /improve reaches Step 4.5 and needs isolated A/B/C/D eval prototype classification, synthetic test cases, mental simulation, and structured diagnostics. tools: Read, Grep, Glob, LS model: sonnet
Skill Verifier
你是 `/improve` 的 surrogate verifier。你的任務不是改寫 skill,而是在資訊隔離下驗證候選改寫是否足以修復 gap,並產出可執行診斷。
Inputs You May Use
只使用呼叫方提供的資料:
- gap 描述與 signal metadata
- Skill Evolution Memory
Alternatives in Testing & QA
- Eval Author — Authoring role for the D365FO agent eval loop catalog 136 ★
- Python Test Engineer — Use this agent when you need to create new tests, fix failing tests, refactor test code, or improve test organ 95 ★
- Sms Assistant Test Writer — Use this agent when the user needs to write or improve unit tests for the SMS assistant functionality, particu 77 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Sim Engineer
Use to write or fix sim.js, the pure simulation logic for the Monty Hall problem (3-door and N-door), after SP
Wh Designer
Design-phase agent. In one session turns " -> " into the risk tier, spec.md with ADRs, evals.md within the tie
Baseline Verifier
Subagent dispatched by do-test to classify test failures as regressions vs pre-existing by running failing tes
Parallel Runner
Isolated worker for an independent unit of work that can run in parallel with other work in the main session.
Fresh Verifier
Fresh-context Opus 5 verifier. Run after a builder reports done. Checks work against the spec, runs tests, rep
Forge Visual Verifier
Perceptual gate for spec [visual] acceptance criteria. Drives Playwright MCP (navigate + take_screenshot + eva
Related Skills
Jiahao
Dual-profile prompt-as-mental-model harness for LLM agents — advisory generator profile in the working agent,
Claude Startup Architecture Console
Unofficial synthetic prototype: Claude Startup Architecture Console for Anthropic's Applied AI Architect, Star
Topoguard
Topology design and a k-of-n corroboration firewall against hallucination spread in multi-agent LLM systems. T