🚁 FlyCLI: Your Drone Under Full Control (and AI)

FlyCLI is a reliable command-line tool for automated interaction with Betaflight flight controllers. Designed for developers, AI agents, and pilots who value stability and automation.

🚀 Installation and Setup

To use flycli as a global command from any folder:

# 1. Clone the repository and enter the folder
git clone https://github.com/archat-hash/FlyCLI.git
cd FlyCLI

# 2. Install dependencies
npm install

# 3. Create a symlink for global access
npm link

Now you can run the tool simply with the flycli command instead of ./index.js.

🗺️ Documentation Navigation

Use the following links for a quick dive into the project:

• 🚀 Quick Start and User Guide
  How to install, scan ports, and execute commands.
• 🏗️ Architecture and Design Decisions
  Description of the C4 model, Clean Architecture, and connection stability mechanisms.
• ⚠️ Important Warnings and Safety
  Must read before using the defaults command.
• 🧪 Testing and Verification
  About running 40+ tests on real hardware and in simulation.

🛠️ Commands (after npm link)

| Command | Description | | --- | --- | | flycli scan | Search for connected flight controllers | | flycli execute <port> 115200 "status" | Execute a CLI command | | flycli health | Complete system checkup (JSON report) |

🔍 "Bottom-Up": How It Actually Works

Unlike "ideal" schemes, real interaction with a flight controller has its nuances:

• Debounce 300ms: We added a synthetic delay after receiving the # prompt because Betaflight often sends data in chunks, and the prompt may appear before the last bytes of the response arrive.
• Handshake MSP: Before entering CLI mode, we perform an MSP request API_VERSION. This is not just a formality — it allows "waking up" the hardware and ensuring the port is ready for data exchange.
• Pure ESM: The project is intentionally written in pure JavaScript (ESM) without TypeScript to minimize build steps and make it as lightweight as possible to run in any Node.js environment.

🦾 AI-Ready

FlyCLI is designed for convenient work with Large Language Models. Use the --json flag to get structured data that your AI agent can instantly analyze.

Created for those who love to fly and code. 🚁💨