Agent-to-agent communication needs no complex protocol, just approvals for secure collaboration

jasonkneen · x · 2026-08-27

The author argues that over-engineering communication protocols for AI agents is unnecessary. Placing agents in a shared channel and requiring manual approval for all requests and responses enables secure and effective collaboration. The author notes that applying human rules to AI often degrades performance and demonstrates a simple implementation.

Original post →

More from coding & agent

coding & agent channel →