Universal design-before-you-build protocol for AI coding agents
<p align="center"> <img src="docs/logo.svg" alt="Preflight" width="100"> </p> <h1 align="center">Preflight</h1> <p align="center"> <strong>Think before you build.</strong> A universal protocol that makes AI coding agents design before they code. <br><br> <a href="https://agentguardhq.github.io/preflight">Website</a> · <a href="PROTOCOL.md">Protocol Spec</a> · <a href="https://github.com/chitinhq/preflight/releases/tag/v1.0.0">v1.0.0</a> </p> Preflight defines 5 mandatory phases every agent completes before writing code: **Orient, Clarify, Approach, Confirm, Execute.** It works with any AI coding agent — no runtime, no API key, one file. ## Why AI coding agents jump straight to implementation. They write code before reading the codebase, skip exploring alternatives, and don't confirm intent. This wastes tokens, produces wrong solutions, and creates rework. Preflight fixes this with a lightweight prompt-based protocol that works across every major agent driver. ## Quick Start ### Auto-detect and install ```bash bash <(curl -fsSL https://raw.githubusercontent.com/chitinhq/preflight/main/install.sh) ``` The installer detects which AI agent drivers you use and copies the appropriate config file. ### Install for a specific driver ```bash bash <(curl -fsSL https://raw.githubusercontent.com/chitinhq/preflight/main/install.sh) --driver claude-code ``` ### Manual install Copy the driver file for your agent into your project: | Driver | Copy this file | To this location | |--------|----------------|------------------| | Claude Code | `drivers/claude-code/preflight.md` | `.cursor/commands/preflight.md` | | Codex (OpenAI) | `drivers/codex/AGENTS.md` | Append to `AGENTS.md` | | Cursor CLI | `drivers/cursor/CLAUDE.md` | Append to `CLAUDE.md` | | Goose | `drivers/goose/.goosehints` | Append to `.goosehints` | | GitHub Copilot | `drivers/cursor/.github/cursor-instructions.md` | Append to `.github/cursor-instructions.md` | | Cursor | `drive
Agent that generates comprehensive documentation, API references, architecture diagrams, and developer onboarding guides from existing code.
Agent configuration for systematic bug investigation that traces issues from error logs through the codebase to root cause with suggested fixes.
Agent for integrating third-party APIs including SDK setup, type generation, error handling, retry logic, and rate limit management.
Cursor's built-in autonomous coding agent that can make multi-file edits, run terminal commands, search the codebase, and iteratively build features with minimal human intervention.
Cloud-based autonomous coding agent that runs in the background on remote sandboxed environments, handling complex multi-step tasks while you continue working.
Cursor's multi-file editing agent within Composer mode that can create, edit, and delete files across your entire project in a single conversation.