An AI agent tells you it did the job. av9n checks that claim against independent evidence — the payment processor, the CI system, the ticketing log — not the agent's own word. Here's what that catches.
Pick a scenario — the agent reported success. See what the evidence says.
Real verification against a live external system
Enter any public GitHub repo and a claim about its latest build. av9n pulls the actual CI run from GitHub — a system the agent can't fake — and checks your claim against it. This is the real thing, live.
Public repo (owner/name)
The agent's claim about the latest build
Verify your own claim
Enter a claim an agent might make, and the independent evidence that should back it. av9n checks whether the evidence actually supports the claim.
The agent's claim
Independent evidence (a system the agent doesn't control)
Check an agent log for contradictions
Paste what your agent claimed it did, and its actual log/transcript. av9n flags contradictions inside the log — like reporting success while the log shows an error. (This is a consistency check; full verification goes further, against outside evidence.)