Show HN: Approv – human approval and a signed audit trail for AI agent actions

Category: infrastructure

Tags: ai-agents, human-in-the-loop, audit-trail

Score: 5.7/10 (Innovation: 5, Technical: 6, Documentation: 5, Utility: 6)

Approv provides a simple API to pause AI agent actions and require human approval via WhatsApp or SMS, with a signed Ed25519 audit trail for tamper-evident records. It addresses the growing need for safe human-in-the-loop control in autonomous agent workflows, combining practical messaging integration with cryptographic verification.

Target audience: backend devs, devops, data engineers

View on Hacker News