codeaholicguy/ai-devkit

The control plane for AI coding agents.

What it solves

AI DevKit provides a unified control plane for developers who use multiple AI coding agents (such as Claude Code, Cursor, and GitHub Copilot). It solves the problem of fragmented setups, where each agent has its own configuration, isolated terminal tabs, and no shared memory or communication paths, often leading to inconsistent behavior and a lack of verifiable progress.

How it works

The tool acts as an operating layer that sits above your existing AI agents. It uses a single configuration file (.ai-devkit.json) to reconcile setup across different tools. It provides a TUI dashboard (agent console) for supervising sessions and a CLI (agent send) for routing prompts and logs to running agents. To prevent context bloat, it uses a local SQLite database for memory retrieval, allowing agents to search for project conventions and decisions only when needed. It also implements a structured dev-lifecycle workflow that forces agents to move through requirements, design, planning, implementation, and testing phases before marking a task as complete.

Who it’s for

Developers and engineering teams who use a variety of AI coding assistants and need a way to coordinate them, maintain shared project rules, and ensure a high standard of engineering rigor (like TDD and verification) across different providers.

Highlights

  • Unified Control Plane: Manage and supervise multiple AI agents from a single console and configuration.
  • Retrieval-Based Memory: Local SQLite storage for project decisions and conventions to reduce prompt bloat.
  • Composable Engineering Skills: Built-in skills like verify, tdd, and structured-debug that enforce professional software engineering workflows.
  • Cross-Agent Communication: Route prompts, logs, and stdin to running agents or even operate agents via Telegram.
  • Agent-Agnostic Setup: Works across a wide range of agents including Claude Code, Gemini CLI, Cursor, and Devin.