See LobsterHoney in Action
Watch a real AI agent get caught, analyzed, and its system prompt extracted — all in real-time.
Live Incident View
CRITICAL
AI_AGENT_MALICIOUS
Confidence: 99%
Threat Score: 98
Attack Timeline
Extracted Intelligence
Operator
--
Model
--
Purpose
--
Tools
--
Extracted System Prompt
Loading...
Try It Yourself
Hit our public demo trap and watch your request appear in real-time below.
curl -X GET /demo/trap
Rate limited to 10 requests per hour. Data resets hourly.
Live Hits
Waiting for hits... Run the curl command above to see your request appear here.
Ready to catch AI agents on YOUR infrastructure?
Deploy honeypot traps in minutes. No code changes required.
Sign Up Free