DEDUPED Hot AI repos, turned into use cases Get 5 a day

How to audit agent tools for poisoning and excess privilege with Skills

Review an agent's tool definitions for poisoning, excess privilege, and deployment blast radius before rollout.

Authorization notice. Skills is dual-use. Only use it on software, sites and systems you own or have explicit permission to test.

Who it is forAn AI platform engineer about to give an agent new tools who needs a review before the agent goes live.

The problemAgent tools can carry poisoned instructions, unsafe parameters, or more privilege than the task needs.

The outcomeA ranked list of tool-surface findings, each tied to the tool definition it came from.

What you need first

  • Written authorization and rules of engagement for the target system
  • A coding agent that reads SKILL.md, such as Claude Code or Codex
  • Supplied evidence for the assessment, such as tool definitions, exported policy, or test-tenant fixtures

Steps

  1. Install the plugin marketplace in Claude Code with /plugin marketplace add zero-lab-ai/skills.
  2. Run /plugin menu and select the agent-security bundle.
  3. Give the agent-surface-audit skill the tool definitions and deployment context you are authorized to review.
  4. Check each finding against the tool's real permissions before changing any configuration.

Prompt for your agent

Using the agent-surface-audit skill from the zero-lab-ai/skills repo, review the tool definitions for <agent-name> that I am authorized to assess. Return findings for poisoning, excess privilege, unsafe parameters, and deployment blast radius, each with the tool definition it cites. Verify that every finding quotes a definition I provided.

Caveat. The Skills support authorized work and do not grant permission or replace your organization's rules of engagement.

Share: X LinkedIn Hacker News Email Copy link

Disclaimer

Deduped lists open-source repositories that belong to other people. We do not own, maintain, audit, or vouch for any of them, and a repo appearing here is not an endorsement or a safety check. The use cases, steps, and prompts are written by AI and may be incomplete, outdated, or just wrong. Where a use case is marked "Seen in the wild", we link to public posts where someone described something similar, but we have not verified those posts and the use case itself is still our AI's own write-up. Use all of this at your own risk. Only run code, commands, or agents on systems and accounts you are authorized to use, and read the code and every command before you run it, especially anything that touches credentials, files, or the network. Nothing here is professional, security, legal, or financial advice. All trademarks and project names belong to their respective owners, and their use here does not imply affiliation. Content is provided as is, without warranties of any kind. Found a mistake, a safety problem, or something you want removed? Email [email protected] and we will look at it promptly. Full disclaimer page