# Pi Security agent skills for Claude Code, Cursor and RevampKey

> Ready-made Pi Security playbooks for AI agents: ingest a Markdown security report into Pi Security, investigate a Pi Security finding with AI, find or prepare remediation guidance in Pi Security. Free SKILL.md files for Claude Code, Cursor and Codex, or one click in RevampKey.

## Skills

- [Use Pi Security safely with AI](https://revampkey.com/marketplace/pi-security/skills/essentials.md): Confirms which Pi tenant you are querying, picks the right lookup tool, keeps results narrow, and asks before changing anything in Pi.
- [Ingest a Markdown security report into Pi Security](https://revampkey.com/marketplace/pi-security/skills/ingest-markdown-report.md): Submits a security report pasted as Markdown or plain text to Pi Security for ingestion. It confirms the tenant and report name, screens the content for secrets, and asks for your explicit yes before the one hosted upload.
- [Investigate a Pi Security finding with AI](https://revampkey.com/marketplace/pi-security/skills/investigate-finding.md): Reads one Pi Security finding by display id, UUID or URL, explains it, and shows any existing remediation plan. It only reads: it never changes anything in Pi.
- [Find or prepare remediation guidance in Pi Security](https://revampkey.com/marketplace/pi-security/skills/remediation-guidance.md): Looks up existing remediation plans for a Pi finding or vulnerable package first, and only prepares a missing package plan after you confirm. It does not change code.
- [Review security posture with Pi Security](https://revampkey.com/marketplace/pi-security/skills/review-security-posture.md): Combines an application's threat model with Gatekeeper pull request and issue state to show where open risk concentrates and which boundary-touching changes have not been reviewed. Read-only, so it asks nothing before running.
- [Start a Pi Security design review](https://revampkey.com/marketplace/pi-security/skills/security-design-review.md): Creates a hosted Pi design review from a Confluence or Notion URL or inline Markdown. It checks the input, confirms the tenant and target, and asks before submitting anything.

## Install

- RevampKey: connect Pi Security at https://revampkey.com/marketplace/pi-security; the skills come with it.
- Any Agent Skills tool, into the current project:

```bash
for s in essentials ingest-markdown-report investigate-finding remediation-guidance review-security-posture security-design-review; do
  curl -fsSL --create-dirs -o .agents/skills/pi-security-$s/SKILL.md https://revampkey.com/marketplace/pi-security/skills/$s/SKILL.md
done
```

### What is a Pi Security agent skill?

A SKILL.md file: a short playbook an AI agent loads when your request matches it. Each Pi Security skill describes one job, which Pi Security tools to use in which order, what to check first and when to stop and ask you.

### Do the Pi Security skills change anything on their own?

3 of the 6 only read. The others tell the agent to show you the change and wait for a yes before it creates, edits or deletes anything in Pi Security.

### Which tools can use these skills?

Any agent that reads the open Agent Skills format: Claude Code, Cursor, Codex, VS Code with Copilot and RevampKey. In RevampKey they are attached when you connect Pi Security; elsewhere you copy the folders into your project or skills directory.

### Do I need the Pi Security MCP server too?

Yes. The skills tell the agent how to do the work; the Pi Security MCP server gives it the tools to do it. The Pi Security MCP setup guide covers every client.

MCP setup guide: https://revampkey.com/marketplace/pi-security.md