When Your AI Agent Passes 2,283 Tests — And Still Fails in Production

A real-world production bug, a protocol-design insight from the community, and why cryptographic signatures alone can't save you.

Two weeks ago, I posted on LinkedIn about OpenWorkProof, a verification protocol for AI agent work. I expected maybe a dozen likes. I did not expect a global community of engineers, protocol designers, and security researchers to spend 72 hours dismantling our assumptions in public — and then rebuilding them into something better.

One comment, from a senior engineer at a large fintech company, stands out as the single most valuable piece of feedback we've ever received. He described a production bug that had survived for months:

The Verifier That Couldn't Fail

Their gateway ran a verification check on every API response. The check was signed, audited, and reported verified: true on every run. For months, nobody questioned it.

When Your AI Agent Passes 2,283 Tests — And Still Fails in Production

Related reading

I built a protocol so AI agents can't just say 'trust me'

When AI Agents Ship Code: A Protocol for Verifiable Execution

Your AI Agent Passed All Tests — Then Failed in Production. Here's the…

The most dangerous line of code your AI agent writes is the test that passes

When Your AI Agent Handles Money, "It Worked" Isn't Good Enough

I Ran 150 Tasks to Test If AI Agents Follow Rules — The Answer Surprised Me