SKILL
agent-security-hardening
Security hardening guidance for AI agent frameworks that process untrusted content, invoke tools, write workspace files, manage runtime identifiers, or handle credentials. Use when building or reviewing an agent runtime, autonomous worker, tool gateway, memory service, or multi-tenant agent deployment. Do not use for general web application security or offensive testing.
Target: Cloud, operations & security. Automatically collected · unverified.
Use with Cursor
A guide for this combination is not available yet, and compatibility has not been verified.
Requirements
- Only the author’s Skill metadata is included. Read the full instructions to check dependencies, licensing and conditions of use.
Choose an AI tool with documented instructions, or save the resource and request a guide after the page loads.
Read the author’s documentation ↗