~ / cases / system-prompt-exfiltration

⌁ System Prompt Exfiltration

Testing whether hidden instructions, policies and system context can be extracted from AI-powered applications.

[docker lab] prompt-leak

Probe prompt disclosure safely with canaries and defensive checks. Run it locally to practice the proof of concept in an isolated environment.

aurora@hack:~$ ls ./cases/system-prompt-exfiltration

System Prompt Exfiltration Through Agentic Workflow Creation

The application featured an AI-powered chat for creating workflows.