Documentation
Set up the control plane that gives your coding agents one config, one console, local-first memory, communication, workflow skills, and verification.
What is AI DevKit?
An overview of AI DevKit, the control plane for AI coding agents.
Getting Started
Set up AI DevKit as a control plane for Claude Code, Codex, Cursor, opencode, and other AI coding agents.
Supported AI Agents & Environments
Compare supported AI coding agents and environments for AI DevKit setup, skills, MCP configuration, and local control-plane workflows.
Development with AI DevKit
Run an agentic software development workflow where AI coding agents plan, implement, test, verify, and review before push.
Debugging with AI DevKit
Learn how to systematically debug issues using AI DevKit's structured debugging workflows
Understanding Existing Code with AI DevKit
Learn how to analyze and document existing codebases using AI DevKit's document-code command
Memory
Give AI coding agents local long-term memory with MCP, CLI commands, and project-scoped knowledge storage.
Skills
Install reusable skills that teach AI coding agents team workflows, coding standards, testing patterns, and domain-specific practices.
Agent Management
Manage and interact with other AI agents running on your system
Agent Setup
Configure AI DevKit once and generate repeatable setup for Claude Code, Codex, Cursor, MCP servers, skills, memory, and workflow docs.
Dev Lifecycle Skill
Make your agent plan before code, verify work, and review changes with a repeatable engineering workflow.
Configuration File
Complete reference for the .ai-devkit.json configuration file used by AI DevKit.
Channel
Connect AI agents with messaging channels like Telegram for remote interaction
Agent Console
Monitor, message, start, rename, kill, and connect running AI agents from one multi-agent terminal UI.
Plugins
Install global AI DevKit plugins from npm and author packages that add CLI commands.
Need Help?
Can't find what you're looking for? Check out our GitHub repository or open an issue.