Submit your agent. We run 50+ adversarial scenarios automatically. Receive cryptographically signed behavioral trajectory in 72 hours. No integration required.
Purchase. Complete payment. You receive a submission link immediately.
02
Submit your agent. Provide your agent's API endpoint or system prompt. No SDK integration, no deployment changes.
03
We run 50+ scenarios. Automated adversarial scenarios execute against your agent. Typically completes within 24-48 hours.
04
Receive signed evidence. Behavioral trajectory with Ed25519 signature, framework mapping, conduct rating, and shareable results page. Delivered to your email within 72 hours of submission.
HTML results page (shareable link, includes badge)
PDF summary (one-pager for procurement conversations)
Framework mapping table (NIST, OWASP, EU AI Act columns)
FAQ
Do I need to deploy anything?
No. You submit your agent's system prompt or API endpoint. We handle the rest. No SDK, no middleware, no code changes.
What if my agent fails scenarios?
The evidence shows exactly what happened. Failure is diagnostic information, not a penalty. Most agents fail at least one category. The value is knowing which one before your client discovers it.
Is this a certification?
No. BotConduct is an observatory, not a certification body. We produce diagnostic evidence. You use it however you need — for compliance, procurement, internal decision-making.
Can I share the results with my clients?
Yes. The results page is designed to be shared. The Ed25519 signature means anyone can verify the results independently without trusting us.
What if I need help interpreting the results?
Upgrade to Guided Assessment (Contact for pricing), which includes a 90-minute expert session and mitigation roadmap.