PwnClaw Security Scan

Test your AI agent for security vulnerabilities using PwnClaw. Runs 50+ attacks (prompt injection, jailbreaks, social engineering, MCP poisoning, and more) and provides fix instructions. Use when your agent needs a security check or hardening.

This skill is internally consistent, but it relies on an external service. Before using: verify the PwnClaw website and the linked GitHub repo yourself; prefer human-mediated/manual testing or a sandboxed agent endpoint for initial runs; never give the service access to sensitive production endpoints or credentials; review any system-prompt rules generated by the service before applying them permanently; and limit scope/privileges when running tests (use staging agents or explicit test tokens). If you need higher assurance, inspect the public GitHub source or run the service in a controlled environment first.

Type: OpenClaw Skill Name: pwnclaw-security-scan Version: 1.0.0 The skill bundle is designed to test an AI agent for security vulnerabilities using the PwnClaw service. The `SKILL.md` instructs the agent to make HTTP GET and POST requests to `https://www.pwnclaw.com/api/test/{token}` to fetch attack prompts and submit its responses. This network activity and data transfer (agent's responses to test prompts) are explicitly aligned with the stated purpose of a security testing tool and do not show evidence of exfiltrating unrelated sensitive data, malicious execution, persistence, or obfuscation. The instructions are clear and do not attempt to prompt-inject the agent into performing actions outside the scope of the security scan.