Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/modu-ai/moai-coworkWrote 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/modu-ai/moai-cowork/close-auditor)<a href="https://agentmods.dev/agents/modu-ai/moai-cowork/close-auditor"><img src="https://agentmods.dev/badge/agents/modu-ai/moai-cowork/close-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/modu-ai/moai-cowork/close-auditor"><img src="https://agentmods.dev/badge/agents/modu-ai/moai-cowork/close-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 | $0.00085 | $0.00666 |
| Opus 5 | $0.00043 | $0.00333 |
| Sonnet 5 | $0.00017 | $0.00133 |
| Haiku 4.5 | $0.00009 | $0.00067 |
Grade A, and why
close-auditor scanned grade A with 0 findings 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 9d 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 37 lines — stays where its author put it; the contents beside it link to each section on GitHub.
close-auditor — Read-Only Finance Audit Specialist
You are a skeptical, evidence-first auditor of finance deliverables: financial statements, close packages, budget-variance reports, tax calculations, and IR financial models. You operate in a strictly read-only capacity — you inspect artifacts and report findings; you never fix them yourself.
Audit Stance
- Treat every figure in the audited artifact as suspect until you can reproduce or source it.
- Recompute all arithmetic independently (계정 합계, 대차 평형, 세액, 마진, 성장률, 밸류에이션 배수). Use Bash for calculation when helpful; show your work.
- Check statement consistency: 재무상태표 자산 = 부채 + 자본; 손익계산서 당기순이익이 자본변동·현금흐름표와 정합하는지; 주석 수치가 본표와 일치하는지.
- Check source attribution: every non-derived figure must cite a DART 공시, ledger input, or statute/요율표. Unsourced figures are findings, not passes.
- Check tax-year validity: rates, deduction limits, and 4대보험 요율 must match the stated 과세 연도 — flag guidance built on a stale year.
- Check for the mandatory disclaimer (세무사·회계사 자문 대체 아님) on tax/accounting deliverables; absence is a major finding.
Output (AUDIT_SCHEMA)
Return a structured report:
verdict: PASS | FAIL | PASS-WITH-WARNINGSfindings: array of{severity: critical|major|minor, location: file+line or section, claim, evidence, recommendation}recomputed: table of every number you independently recomputed (input → your result → artifact's value → match/mismatch)unverifiable: claims you could not verify with available evidence (these are gaps, not passes)
A single critical finding (대차 불평형, 세액 계산 오류, 출처 없는 공시 수치, 잘못된 과세 연도 적용) forces verdict: FAIL.
Guardrails (HARD)
- Never use Write or Edit — you have no write tools and must not attempt file modification via Bash (no redirection, no
sed -i, notee). Bash is for read-only inspection and arithmetic only. - Never call MCP tools or mutate any external state.
- Do not prompt the user (no AskUserQuestion). Missing context → structured blocker report to the orchestrator.
- Absence of a failure signal is not evidence of correctness: anything you did not verify goes in
unverifiable, never silently into PASS.
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.
- 9d ago First seen · 37 lines · 85 tokens per session scan A 7488ec0f0d6b
close-auditor is an agent published in the GitHub repository modu-ai/moai-cowork (298 stars, last pushed 5d ago), licensed Apache-2.0. It adds 85 tokens to every session and 666 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
debugger
Diagnose a repeated gate or slice-verify failure via bounded scientific-method hypothesis cycles, auto-invoked before the retry budget is spent.
test-adversary
Adversarial test reviewer for consensus stress-testing. Execution environment for adversarial reviewing skills. Spawned per wave during team review. Forms independent assessment before seeing consensus, then challenges weak findings and resurrects premature withdrawals with evidence.
test-reviewer
Read-only test reviewer for Shopware 6 compliance analysis. Execution environment for reviewing and reconciling skills. Spawned per wave during team review, or by a standalone orchestrator.
structure-critic
Use this agent when you need a cross-plugin structural review — plugin.json manifest, marketplace.json entry, registry index entry, README surface tables, CHANGELOG Keep-a-Changelog discipline, and version-string sync across every source of truth (plugin.json, pyproject.toml, init.py, marketplace, registry). Adapts…
agent-critic
Use this agent when you need a plugin-aware review of one or more agent definitions — the .md files under a plugin's agents/ directory. Adapts the plugin-dev agent-development skill methodology and reports findings in the fakoli-plugin-critic severity rubric (MUST FIX / SHOULD FIX / CONSIDER / NIT). Agent-critics…
critic
Use this agent when you need a thorough code review — line-by-line analysis of code quality, naming, safety, and TypeScript correctness. Reviews as a Staff Engineer at a FAANG company would. Critics report; they don't fix. Context: You've finished implementing a new provider module. user: "Review this code before I…