# codeaholicguy/ai-devkit

The control plane for AI coding agents.

Repository: https://github.com/codeaholicguy/ai-devkit
Canonical: https://ross.abutalabs.com/products/ai-devkit
Homepage: https://ai-devkit.com
Language: TypeScript
License Family: other
Topics: ai, claude-code, antigravity, agent-skills, agent-framework, ai-agents, ai-coding, cli, codex-cli, coding-agents, cursor-ai, developer-tools, gemini-cli, mcp, memory, software-engineering, workflow-automation, opencode, pi
Last push: 2026-08-26T17:52:07+00:00

## Health v2 (maintenance only)
Score: 80/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 23
- inputs: {"age_days": 323, "days_push": 7, "days_rel": 10, "gap_med": 1.5, "n_releases_24m": 63}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1597, forks 244 (observed 2026-08-28T04:05:09.299014+00:00)

## What it is
AI DevKit is a local-first control plane for AI coding agents such as Claude Code, Codex CLI, Gemini CLI, opencode, and Cursor. It provides one shared configuration file, a TUI console for supervising running agents, cross-agent messaging, SQLite-backed memory retrieval exposed via MCP, and composable engineering workflow skills.

## Use cases
- manage multiple ai coding agents from one console
- share one config across claude code and cursor
- give coding agents long-term project memory
- send prompts and logs to a running ai agent
- make ai agents follow a repeatable dev workflow
- stop agents from forgetting project conventions
- orchestrate multi-agent coding sessions locally

## When to choose
- you run more than one AI coding agent and lack a shared setup or control surface
- you want local, searchable memory for agent decisions and conventions without bloating prompts
- you need to route prompts, logs, or test output to running agent sessions from scripts or a TUI
- you want agents to plan, verify, and review work with a structured engineering lifecycle

## When to avoid
- you use a single AI assistant and don't need multi-agent orchestration
- you want a cloud-hosted or team-shared agent management platform rather than a local-first CLI
- you need a framework for building custom AI agents from scratch rather than managing existing coding agents
- your environment doesn't run Node.js 20+ or none of the supported coding agents

## Facets
- artifact type: cli-tool
- maturity: active
- function: agent-framework, mcp, cli, developer-tools, workflow-automation, chat-interface, plugin-system
- domain: developer-tools, large-language-models, apis
- platform: cli, cross-platform, windows
- tags: ai-coding-agents, agent-orchestration, control-plane, agent-memory, multi-agent, claude-code, codex-cli, gemini-cli, cursor, agent-console, dev-lifecycle, local-first, sqlite, agent-skills, tui, ai-agents, command-line, automation, nodejs, macos, linux

## Member repositories
- codeaholicguy/ai-devkit (main) score 80

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:09.299014+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:53:59.281982+00:00, confidence not recorded.
  - readme: https://github.com/codeaholicguy/ai-devkit (fetched 2026-08-28T04:05:09.299014+00:00, sha 4c710ec63c88)
  - homepage: https://ai-devkit.com (fetched 2026-08-29T11:25:03.028699+00:00, sha 417176da6354)
  - site_page: https://ai-devkit.com/docs (fetched 2026-08-29T11:25:03.038094+00:00, sha f1e1390d335c)
  - site_page: https://ai-devkit.com/docs/1-getting-started (fetched 2026-08-29T11:25:03.056760+00:00, sha 5d69557dc11f)
  - registry_npm: https://registry.npmjs.org/ai-devkit (fetched 2026-08-29T11:25:03.060733+00:00, sha 4e7a02e4e873)
  - site_page: https://ai-devkit.com/faq (fetched 2026-08-29T11:25:03.040574+00:00, sha 58055214d0e2)
- Data as of 2026-08-30T08:39:29.467469+00:00.