A credential proxy for AI agents. They never see the secret.
TAP sits between your agents and the APIs they call. The agent sends a placeholder; the proxy injects the real credential server-side, enforces policy, and gates risky actions behind human approval. Your keys never touch the model.
POST /forward cred: <CREDENTIAL:stripe>
→ 202 · awaiting human approval✓ approved · key injected server-side
You're on the list — we'll be in touch with early access.