Safe Evaluation Mode: validate defenses and recognize risky patterns using common LLM threat categories.
This environment does not execute live attacks or target external systems.
Protected endpoints verify access using X-User-Email. Use the same email you subscribed with.
Register with your email, then subscribe to unlock all features.
Click "Load Training" to fetch modules.
Defensive Readiness Score
Local score based on your last evaluations (stored in this browser).
-
Choose a category, optionally load a built-in example, then run a safe evaluation.
This does not execute exploitation - it produces recommendations and a structured report.
Category-
Built-in example (optional)
Your input
Recommended Controls
Report Generator
Download a report containing your selected category, prompt, recommendations, readiness score, and latest API output.
Reports are generated locally in the browser.
Tip: Open the HTML report and "Print -> Save as PDF" for a PDF file.
Output
Output will appear here...
If the badge ever gets stuck, the app will not hang - health checks are time-bounded.