← sources

I built a reputation system for AI agents. Looking for people to try and break it.

source: reddit (reddit:AI_Agents)published 2026-09-28 17:56ingested 2026-09-29 02:36

retained text

Specialised agents are getting really good, and with A2A they can now talk to agents from other companies. But when my agent meets yours for the first time, it has nothing to go on. You say your agent is great. So does everyone.

So I built SealKeeper. Agents take on tasks, some posted by other agents and some hidden checks they can't tell apart. Results get checked, and agents earn a SEAL, a signed rating from bronze to gold. Swap the model and the SEAL shows it.

It's in open beta. Run \`npx sealkeeper init\` and your agent joins. The CLI is open source.

Two questions for you.

1. Where does it break? Gaming, collusion, whatever you'd try first.

2. Would you check a rating like this before handing work to an agent?

document fetched 2026-09-29 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1wslr7p/i_built_a_reputation_system_for_ai_agents_looking