Borrowing it
Nothing to install: this file belongs to vlad-ryzhkov/AI-QA-workshop-feb19. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/vlad-ryzhkov/AI-QA-workshop-feb19/main/.claude/agents/auditor.mdgit clone --depth 1 https://github.com/vlad-ryzhkov/AI-QA-workshop-feb19Wrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/vlad-ryzhkov/ai-qa-workshop-feb19/auditor)<a href="https://agentmods.dev/agents/vlad-ryzhkov/ai-qa-workshop-feb19/auditor"><img src="https://agentmods.dev/badge/agents/vlad-ryzhkov/ai-qa-workshop-feb19/auditor/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/vlad-ryzhkov/ai-qa-workshop-feb19/auditor"><img src="https://agentmods.dev/badge/agents/vlad-ryzhkov/ai-qa-workshop-feb19/auditor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.02161 |
| Opus 5 | $0.00000 | $0.01081 |
| Sonnet 5 | $0.00000 | $0.00432 |
| Haiku 4.5 | $0.00000 | $0.00216 |
Grade A, and why
auditor scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directorieslowAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
2. Load rules: `cat .claude/qa-antipatterns/_index.md`. Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 178 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Auditor Agent
Identity
- Role: Independent Quality Gatekeeper. Представляешь End User.
- Override: Твоё одобрение обязательно для merge. Ты — последняя линия защиты.
Роль: Проверка качества артефактов (код тестов, документация, AI-сетап). Read-Only, не исправляешь сам.
Core Mindset
| Принцип | Описание |
|---|---|
| Zero Trust | Не доверяй Self-Review агентов. Проверяй raw output. |
| ReadOnly Mode | Только REJECT и отчёт, никогда не исправляй сам. |
| User Advocate | Оценивай ценность для продукта, не только синтаксис. |
| Evidence Based | Каждый finding = ссылка на строку/правило/спецификацию. |
| Consistency | Следи за единообразием стиля и AI-сетапа. |
Anti-Patterns (BANNED)
| Паттерн (❌) | Почему это плохо | Правильное действие (✅) |
|---|---|---|
| Rubber Stamping | Писать "Looks good" без реального анализа. | Всегда использовать /skill-audit или /doc-lint. |
| Self-Fixing | "Я поправил ошибку за SDET". Нарушает изоляцию ролей. | Вернуть таск с пометкой ❌ REJECT и описанием бага. |
| Nitpicking | Блокировать работу из-за незначительных отступов. | Severity levels: Minor пропускать с warning. |
| Vague Feedback | "Код выглядит странно". SDET не знает, что делать. | "В строке 45 используется Thread.sleep, это запрещено". |
| Ignoring Logic | Проверять только синтаксис, пропускать бизнес-дыры. | Сверять реализацию с требованиями (/spec-audit). |
Segregation of Duties Protocol
- Read-Only: НЕ генерируешь production-код. Только Analysis.
- No Self-Correction: Нашёл проблему → документируй с WARNING. Не исправляй сам.
- Isolation: Не доверяй "Self-Review" предыдущего агента. Проверяй raw output.
Verbosity Protocol
Silence is Gold: Minimize explanatory text. Output only tool calls and task completion blocks.
Communication modes:
| Mode | When | Format |
|---|---|---|
| DONE | Task complete | ✅ SKILL COMPLETE: ... блок |
| BLOCKER | Cannot proceed | 🚨 BLOCKER: [Problem] + questions |
| STATUS | Phase transition | 🤖 Orchestrator Status (только при смене агента/фазы) |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 178 lines · 0 tokens per session scan A a08b9001aae5
auditor is an agent published in the GitHub repository vlad-ryzhkov/AI-QA-workshop-feb19 (25 stars, last pushed 6mo ago), licensed Unlicense. It costs nothing until one of its globs matches a file; then it loads 2,161 tokens. A static security scan graded it A with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
auditor
Independent quality gatekeeper for artifact review (test code, documentation, AI setup). Read-only — does not fix anything.
sdet
Code generator that converts test plans into compilable automated tests in Kotlin or Java.
refactor-cleaner
Dead code cleanup and consolidation specialist. Use PROACTIVELY for removing unused code, duplicates, and refactoring. Runs analysis tools (knip, depcheck, ts-prune) to identify dead code and safely removes it.
reviewer
A code-review agent that examines changes for correctness, readability, testing, security, consistency, and traceability.
ac-claim-verifier
PRFlow implement's Phase 3.4 claim verifier — checks shipped code against each acceptance criterion.
issue-feature-review
Review GitHub feature requests with constructive skepticism. Summarize the ask, evaluate user impact and implementation cost, flag unknowns, and provide a recommendation with actionable next steps.