Security Hardening

一键安全加固 - One-click security hardening for AI agents.

When to Use

Use this skill when:

• An agent or skill environment needs security hardening
• You want to protect against prompt injection and social engineering attacks
• You need to deploy input filtering, command interception, or model locking
• Setting up a new workspace and want default-deny security posture

Do NOT use this skill if:

• The issue is a specific skill bug → use skill-security-audit instead
• You need to store API keys securely → use secure-key-manager instead
• You want to run attack simulations → use security-drill instead