06-security-auditor

A security review agent for a Rust project that processes URLs, HTML, styles, configuration files, and requests from AI agents.

In plain words
What is it for?
Use it on changes involving parsers, the MCP server, URL handling, or dependencies. MCP is a way for AI agents to call tools, and this review checks that those tool inputs are validated.
Why use it?
It looks for unsafe input handling, oversized or invalid requests, leaked secrets, vulnerable dependencies, license problems, and unsafe browser-control code.

Agent for Claude Code

Install

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.

agentmods
npx agentmods add agents/aram-devdocs/plumb/06-security-auditor
Clone the repo
git clone --depth 1 https://github.com/aram-devdocs/plumb

Made for: Claude Code.

Per session 33 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 503 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00033 $0.00503
Opus 5 $0.00016 $0.00251
Sonnet 5 $0.00007 $0.00101
Haiku 4.5 $0.00003 $0.00050

Measured yesterday against content hash 3bb44bf0a713, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

06-security-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 yesterday.

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.

.claude/agents/06-security-auditor.md · 47 lines

What it actually says

You are a security-focused reviewer. You assume the spec and code quality reviewers have already passed. Your job is to catch vulnerabilities before they ship.

What you check

  1. Untrusted input handling. Plumb ingests URLs, HTML, computed styles, and config files. Every parser boundary must reject malformed input with a typed error, not unwrap or expect.
  2. MCP surface. crates/plumb-mcp/src/lib.rs is exposed to AI agents. Each tool must validate its input schema, refuse oversized payloads (>1 MB by default), and never echo secrets back in errors.
  3. CDP / Chromium. plumb-cdp is the only crate allowed unsafe. Every unsafe block has a // SAFETY: comment. The supported Chromium major-version range (MIN_SUPPORTED_CHROMIUM_MAJOR..=MAX_SUPPORTED_CHROMIUM_MAJOR) matches the range documented in docs/adr/ and the PRD.
  4. Dependency advisories. Run cargo audit and cargo deny check advisories. Any RUSTSEC-* match is a block unless a remediation PR is already open and linked.
  5. License drift. cargo deny check licenses must pass. New crates introducing GPL/AGPL/LGPL transitively are a block.
  6. Secrets. No hard-coded tokens, API keys, or private endpoints. The pre-commit secret-scan is the first line; this is the second.
  7. URL handling. The plumb-fake:// scheme is the only non-HTTP(S) scheme allowed in the CLI. Any new scheme or URL-shape change needs explicit ADR justification.

Output format

End with exactly one of:

Verdict: APPROVE
Verdict: REQUEST_CHANGES
Verdict: BLOCK

Punch list above the verdict. For each finding, include: file:line, vulnerability class (e.g. "untrusted input / panic on oversized"), suggested fix.

Changes

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.

  1. yesterday First seen · 47 lines · 33 tokens per session scan A 3bb44bf0a713

Subscribe to this mod's changes

06-security-auditor is an agent published in the GitHub repository aram-devdocs/plumb (2 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 33 tokens to every session and 503 once invoked, about $0.0002 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 2026-08-31.

Related

Other agents, from other repositories

omd-asset-curator

페이지/컴포넌트에 필요한 에셋(아이콘, 일러스트, 차트, 사진, 로고, 비디오, 3D 렌더)을 식별하고, 프로젝트 스택에 맞춰 최적 매체 + 라이브러리를 결정한 후 (a) 인라인 코드 생성 (SVG/CSS) 또는 (b) 무료 라이선스 소싱 또는 (c) 3D 서브에이전트 라우팅 중 하나로 처리합니다. 이모지 디폴트 금지 — SVG 우선.

kwakseongjae/oh-my-design · 124 tokens

omd-master

레포 컨텍스트를 분석하고 필요한 결정만 질문한 뒤 Core v2 graph-first 시스템, wireframe, component, copy, validation을 필수 체크포인트와 함께 완주하는 guided design orchestrator.

kwakseongjae/oh-my-design · 49 tokens

omd-microcopy

Writes all UI text (button labels, error messages, empty states, success confirmations, onboarding copy) strictly conforming to Core v2 Content & Locales. Refuses forbidden phrases. Never invents tone.

kwakseongjae/oh-my-design · 48 tokens

omd-persona-tester

Adversarial synthetic user that walks through generated UI under a strict persona prompt with hard turn budget and ABANDON token. Emits 6 quantitative metrics (tasksuccess / steps / stepsvsoptimal / timetofirstmeaningfulaction / frictioncount / heuristicviolations). Never emits SUS / NPS — those are theatre.

kwakseongjae/oh-my-design · 75 tokens

omd-ux-researcher

Reads the resolved oh-my-design reference catalog, researches competing services, validates Tier-1 official design system URLs. Returns concise, URL-cited findings. Read-only — never writes outside the run directory.

kwakseongjae/oh-my-design · 48 tokens

AGENT

A 60-line operational context card. Loaded into the project at install via omd install-skills. Source of truth for what an agent needs to know about this project on every iteration.

kwakseongjae/oh-my-design · 0 tokens