The standard for AI-assisted software development.
AI DevKit is an open-source ecosystem designed to bridge the gap between human intent and AI execution. It provides a suite of tools that standardize how AI agents interact with codebases, manage context, and execute development workflows.
The capabilities of AI coding agents are growing exponentially, but they often struggle with:
- Context: Losing track of broad project requirements and architectural decisions.
- Consistency: Generating code that doesn't follow project-specific patterns.
- Process: coding without a structured plan, leading to "spaghetti code".
AI DevKit solves this by enforcing a Phase-Based Development lifecycle—ensuring agents follow the same rigorous engineering standards as senior developers: Requirements → Design → Planning → Implementation → Testing.
AI DevKit is a monorepo containing multiple specialized tools:
The command-line interface that orchestrates the development lifecycle.
- Scaffold structured documentation (Requirements, Design, Plan).
- Configure environments for Cursor, Claude Code, and other agents.
- Manage project state and development phases.
A lightweight MCP-based memory service that gives agents persistent, searchable long-term memory.
- Store architectural decisions and coding patterns.
- Search documentation and knowledge using full-text search.
- Share context across different agents and sessions.
To start using the core toolkit in your project instantly:
npx ai-devkit initThis will launch the interactive setup wizard to configure your project for AI-assisted development.
We welcome contributions! Whether you're building a new agent integration, adding a memory adapter, or improving our templates.
-
Clone the repository:
git clone https://github.com/Codeaholicguy/ai-devkit.git cd ai-devkit -
Install dependencies:
npm install
-
Build packages:
npm run build
We use a monorepo structure. Ensure all changes are covered by tests before submitting a PR.
MIT