Relaylens probe to verify your API relay isn't cheating you

A lot of us run self-hosted gateways or use third-party relays to access OpenAI/Anthropic models. The problem is, you're trusting the relay to actually serve the model you're paying for. I built Relaylens to solve that. It's a deterministic probe that checks model authenticity using the provider's native signature mechanisms (the thinking/response signatures that a proxy can't fake).