When this page loaded, it ran all 150 prompts from the project's two red-team test sets through the same engine you are about to use. The first set has 50 secrets and personal-data cases, including traps that should not trip it, like Git hashes and invalid card numbers. The second has 100 prompt-injection cases, including innocent sentences that use words like "ignore" and "developer mode". The browser engine is also checked against the original Python engine, finding by finding, on every commit.
Gatekeep checks every prompt on its way to the Claude API. Pick one, send it, and see what happens. It all runs in your browser.
Every decision gets a row. The prompt text itself never does.
Same columns as the proxy's real audit table. Gatekeep keeps a SHA-256 fingerprint of each prompt, what kinds of things it found, where the request went, and how long the check took. The raw prompt is never written down, even here.