agentleFS
Sign inStart free
docxologydocxology/codomyrmex/.agents/skills/red-team/AGENTS.md

codomyrmex / red-team

AGENTS.md10 starsChanged 38 days agoMITView on GitHub

Portable repository-scoped skill: Adversarially test a design, implementation, agent workflow, or tool surface for realistic failure and abuse paths within authorized scope. Defined entirely by SKILL.md. - SKILL.md — canonical definition: frontmatter (name: red-team, description), a five-step workflow (assets, trust boundaries, and out-of-scope actions → enumerate abuse and failure paths → exercise the highest-risk paths with real components → record evidence and severity → re-test mitigations), and the reporting rule (findings first, ordered by risk). - README.md — human signpost.

The file

We split this AGENTS.md into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use for security, reliability, permission, and prompt-injection review.

    line 32
  2. Distinctive:

    Never weaken the authorized-scope boundary or the ban on destructive actions.

    line 35
  3. Distinctive:

    Keep the distinction between confirmed findings and hypotheses, including the "not exploitable under current scope" outcome.

    line 37

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

AGENTS.md — codomyrmex/.agents/skills/red-team

Purpose

Portable repository-scoped skill: Adversarially test a design, implementation, agent workflow, or tool surface for realistic failure and abuse paths within authorized scope. Defined entirely by SKILL.md.

Key Files

  • SKILL.md — canonical definition: frontmatter (name: red-team, description), a five-step workflow (assets, trust boundaries, and out-of-scope actions → enumerate abuse and failure paths → exercise the highest-risk paths with real components → record evidence and severity → re-test mitigations), and the reporting rule (findings first, ordered by risk).
  • README.md — human signpost.

Dependencies

  • Claude Code wrapper: .claude/skills/red-team/SKILL.md only points back to this file.
  • Third step of the composition order in docs/agents/agent-interoperability.md, after systems-thinking and before mcp-tool-use.
  • Its "real components over mocks" step matches the repository zero-mock policy in ../../../AGENTS.md.
  • Parent: ../AGENTS.md

Development Guidelines

  • Use for security, reliability, permission, and prompt-injection review.
  • This is the canonical copy; runtime wrappers such as .claude/skills/red-team/ stay pointers — change the workflow here.
  • Never weaken the authorized-scope boundary or the ban on destructive actions; those lines are the skill's safety contract.
  • Keep the distinction between confirmed findings and hypotheses, including the "not exploitable under current scope" outcome.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free