Agent Guardrails

Agent Guardrails is an open-source, local safety and validation layer for AI coding agents. Designed for developers and engineering teams, it refines vague programming requests into well-bounded tasks, enforces coding boundaries, validates relevant changes, and outputs reviewer-friendly summaries before code merges. It works alongside popular AI code agents (e.g., Claude Code, Codex, Cursor, Gemini, OpenCode), providing scope, validation, and risk checks to maintain codebase safety and trust in the AI-driven development process.