audit-plugins

A command that scans a Claude Code plugin marketplace and compares its agents, commands, skills, and configuration files with Claude Code's official guidance.

In plain words
What is it for?
It audits plugin structure and produces a scored report covering manifests, agents, commands, skills, and MCP server configurations.
Why use it?
It helps identify missing requirements and other issues in plugin definitions, with findings organized by priority.

Command 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 commands/sgaunet/claude-plugins/audit-plugins
Clone the repo
git clone --depth 1 https://github.com/sgaunet/claude-plugins

Made for: Claude Code.

Per session 11 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,878 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 $0.00011 $0.01878
Opus 5 $0.00005 $0.00939
Sonnet 5 $0.00002 $0.00376
Haiku 4.5 $0.00001 $0.00188

Measured 2d ago against content hash 631c6f451408, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

audit-plugins 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 2d 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.

.claude/commands/audit-plugins.md · 166 lines

How it starts

The opening of the file, as written. The whole thing — 166 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Audit Plugins Command

Scan all plugin definitions in this marketplace and compare them against official Claude Code documentation. Produce a scored audit report with prioritized findings.

Process

Step 1: Discover Plugin Structure

Use Glob to enumerate all files to audit:

  • plugins/*/agents/*.md — agent definitions
  • plugins/*/commands/*.md — command definitions
  • plugins/*/skills/*/SKILL.md — skill definitions
  • plugins/*/.claude-plugin/plugin.json — plugin manifests
  • .claude-plugin/marketplace.json — marketplace manifest
  • plugins/*/.mcp.json — MCP server configurations

Count totals (plugins, agents, commands, skills) for the report summary.

Step 2: Launch 3 Parallel Sonnet Agents

Launch all 3 agents concurrently using the Task tool. Each agent must first fetch the relevant official documentation via WebFetch before performing its analysis.


Agent 1: Agent & Subagent Auditor (general-purpose, model: sonnet)

Prompt the agent to:

  1. Fetch official docs using WebFetch:
    • https://docs.claude.com/en/docs/claude-code/plugins-reference — extract agent frontmatter fields, required vs optional, content structure
    • https://docs.claude.com/en/docs/claude-code/sub-agents — extract sub-agent best practices, model selection guidance
  2. Read all files matching plugins/*/agents/*.md
  3. For each agent file, check:
    • Frontmatter fields: name (required), description (required), model (required, valid values: sonnet/opus/haiku), allowed-tools (recommended), color (optional), context (optional), permissionMode (optional), capabilities (optional), skills (optional)
    • Description quality: Should clearly state WHEN to use the agent (shown in Claude Code UI)
    • Model appropriateness: sonnet for code/speed tasks, opus for complex analysis/docs, haiku for fast/simple tasks
    • Content structure: Should include role statement, proactive triggers, core capabilities, implementation patterns, deliverables
    • Proactive triggers: File patterns and keyword triggers for automatic activation
    • Tool restrictions: allowed-tools should be scoped appropriately (not overly broad)
  4. Return findings with severity levels and file references

Read the full file on GitHub · 166 lines

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. 2d ago First seen · 166 lines · 11 tokens per session scan A 631c6f451408

Subscribe to this mod's changes

audit-plugins is a command published in the GitHub repository sgaunet/claude-plugins (16 stars, last pushed 7d ago), licensed MIT. It adds 11 tokens to every session and 1,878 once invoked, about $0.0001 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.