Show HN: Reright.it – Human approval for agent-written text

Category: security

Tags: human-in-the-loop, ai-safety, approval-workflow, agent-tools, mcp

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

Reright provides human approval gates for agent-written text, blocking unapproved sends across git, email, and browsers. It enforces a 24-hour approval window via hooks and an MCP server, though bypasses exist. The project addresses a growing need for oversight in AI-assisted workflows.

Target audience: developers using AI coding agents, devops engineers, security-conscious teams

Repository: https://reright.it/

View on Hacker News