# AGENT SECURITY SCANNER
Test your AI agent for indirect prompt injection in 3 steps.
────────────────────────────────────────────────────────────────
# STEP 1 — GENERATE TEST URL
We'll create a honeypot page with hidden prompt injection payloads.
We'll send your security report to this email.
# STEP 2 — LET YOUR AGENT READ THE URL
Copy this prompt and send it to your agent:
Wait for your agent to respond, then paste its full output in the next step.
# STEP 3 — ANALYZE AGENT OUTPUT
Paste your agent's full response below. Analysis runs locally in your browser.
────────────────────────────────────────────────────────────────
# SERVER-SIDE DETECTIONS
These are triggered when the agent actually executes commands or visits URLs from the page.
────────────────────────────────────────────────────────────────
scanning...