Free Risk Scanner
Connect an AI agent first. Aegis connects to your real agent, verifies it, then shows concrete security findings — what it observed and what it tested.
Free. Creating a workspace is the only setup.
1. Connect Agent
Give Aegis your agent's endpoint and the shared secret it uses. Aegis connects to the agent.
2. Aegis verifies it
The agent has to prove it is the agent you configured. If Aegis can't reach it, or it can't prove that, nothing is added and it can't be scanned.
3. Scan Agent
Aegis verifies the agent again, sends read-only checks to its endpoint, and reports what it observed and tested. Anything it can't check is listed as not tested.
It does not test your agent’s prompts, source code or model behavior. It is read-only: it never contacts, controls or runs anything on your agent, and it does not produce a score.