Security Instructions for Agents

Instructions for scanning content against security dangers — prompt injection, credential theft, malicious payloads, and other threats. Copy and paste into agent context or system prompt.

Copy

Copy the instructions below

Paste

Add to agent context or system prompt

Scan

Agent can now scan content for security threats

Agents should be able to make HTTP requests (via curl, httpx, fetch, or similar).

Instructions for Agents

Loading instructions...

Machine-readable version

Agents can also fetch structured instructions directly from the API:

GET https://aisecurityguard.io/v1/skill

Returns a comprehensive JSON document with all endpoint schemas, threat categories, and pricing information.