Adversarial ML evaluation lab — choose how you want to drive it.
Interactive UI for evasion, inference, extraction, poisoning, and the Adversarial Penetration Test.
Interactive API reference — try any endpoint from the browser.
Open in a new tab — pick "One-click into Postman desktop", import by URL, or download the JSON.
Download ready-to-upload params.json files for every attack — drop one into the params_file slot in Swagger, no typing.
Full reference for endpoints, attacks, parameters, and examples. Click to open · download as Markdown from the top of the page.
Liveness endpoint — should return {"ok":true}.