gpt55-critical-auditor

gpt55-critical-auditor is a skill for Claude Code from richfrem/agent-plugins-skills. It costs 41 tokens per session (473 once invoked), scanned A, original, MIT.

An adversarial audit tool for looking for ways agent plugins, skills, and orchestration can break runtime rules, mutate files unsafely, or bypass tests.

In plain words
What is it for?
Use it to examine execution enforcement, mutation safety, evaluation coverage, baseline integrity, and isolation between plugins.
Why use it?
It searches for failure paths and missing safeguards that a normal compliance review might overlook.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: mentions subagents.

Part of the agent-agentic-os plugin — 26 skills, 4 commands, 6 agents, 3 hooks shipped together

Good fit Use it to examine execution enforcement, mutation safety, evaluation coverage, baseline integrity, and isolation between plugins.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/richfrem/agent-plugins-skills/gpt55-critical-auditor
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.

Any agent
npx skills add richfrem/agent-plugins-skills --skill gpt55-critical-auditor
Clone the repo
git clone --depth 1 https://github.com/richfrem/agent-plugins-skills

Made for: Claude Code.

Or install agent-agentic-os, the plugin that ships this one along with the rest of its 26 skills, 4 commands, 6 agents, 3 hooks.

Wrote 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.

agentmods badge for gpt55-critical-auditor

README.md
[![agentmods](https://agentmods.dev/badge/skills/richfrem/agent-plugins-skills/gpt55-critical-auditor/github.svg)](https://agentmods.dev/skills/richfrem/agent-plugins-skills/gpt55-critical-auditor)
Your own site
<a href="https://agentmods.dev/skills/richfrem/agent-plugins-skills/gpt55-critical-auditor"><img src="https://agentmods.dev/badge/skills/richfrem/agent-plugins-skills/gpt55-critical-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.

agentmods 80×15 button for gpt55-critical-auditor

Your own site · 80×15
<a href="https://agentmods.dev/skills/richfrem/agent-plugins-skills/gpt55-critical-auditor"><img src="https://agentmods.dev/badge/skills/richfrem/agent-plugins-skills/gpt55-critical-auditor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 41 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 473 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00041 $0.00473
Opus 5 $0.00020 $0.00236
Sonnet 5 $0.00008 $0.00095
Haiku 4.5 $0.00004 $0.00047

Measured 11d ago against content hash b5c1e3c932b7, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

gpt55-critical-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 11d 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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

plugins/agent-agentic-os/skills/gpt55-critical-auditor/SKILL.md · 91 lines

What it actually says

Purpose

This skill performs a deep, adversarial system audit designed to break:

  • runtime enforcement guarantees
  • mutation safety assumptions
  • eval coverage completeness
  • plugin isolation boundaries

This is NOT a compliance review.

This is a failure-seeking audit.


Audit Mandate

You must assume:

  • The system is incorrect until proven otherwise
  • Enforcement can be bypassed unless proven impossible
  • Evals are incomplete
  • Plugins will attempt to violate boundaries
  • Mutation logic will corrupt state unless isolated

Required Focus Areas

1. Execution Enforcement

  • Can HALT be bypassed?
  • Are there code paths where violations do not interrupt execution?
  • Are exceptions truly global?

2. Mutation Integrity

  • Can partial mutations persist?
  • Are sandbox boundaries leak-proof?
  • Can state escape temp/sandbox/?

3. Eval Authority

  • What behaviors are NOT covered by evals?
  • Can mutations exploit blind spots?
  • Can regression be reclassified as acceptable?

4. Baseline Integrity

  • Can baselines be indirectly manipulated?
  • Are diffs actually verified or just declared?

5. Plugin Isolation

  • Do any skills:
    • chain multiple actions?
    • make decisions?
    • orchestrate workflows?

6. Orchestration Leakage

  • Are there hidden multi-step flows inside skills?
  • Are sub-agents behaving like orchestrators?

7. Debt System Exploits

  • Can debt accumulate without blocking progress?
  • Can agents route around debt enforcement?

Output Requirements

For every issue:

  • ID
  • Severity (P0/P1/P2)
  • Exploit scenario (MANDATORY)
  • Exact failure path
  • Why enforcement fails
  • System impact
  • Recommended fix
  • Can agent exploit automatically? (YES/NO)

Critical Rule

If you cannot prove a guarantee is enforced:

→ It must be treated as NOT enforced

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 11d ago First seen · 91 lines · 41 tokens per session scan A b5c1e3c932b7

Subscribe to this mod's changes

gpt55-critical-auditor is a skill published in the GitHub repository richfrem/agent-plugins-skills (6 stars, last pushed yesterday), licensed MIT. It adds 41 tokens to every session and 473 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.