Show HN: Burrow – Runtime Security for AI Agents

Category: security

Tags: security, ai-agents, runtime-monitoring, policy-enforcement

Score: 7.3/10 (Innovation: 7, Technical: 7, Documentation: 5, Utility: 8)

Burrow is a runtime security layer for AI agents that intercepts and evaluates tool calls against plain-language policies to prevent credential theft, data exfiltration, and unauthorized actions. It's interesting because it addresses the emerging security gap where traditional monitoring tools don't understand AI agent behavior chains, and it works across popular frameworks like Claude Code, Cursor, and LangChain.

Target audience: devops, security-engineers, ai-developers

View on Hacker News