# xintaofei/codeg

Collaborative multi-agent AI coding workspace: aggregate sessions from Claude Code, Codex, OpenCode, Pi, Grok Build, etc. Desktop app, self-hosted server, or Docker.

Repository: https://github.com/xintaofei/codeg
Canonical: https://ross.abutalabs.com/products/codeg
Homepage: https://docs.codeg.app
Language: Rust
License: Apache-2.0
License Family: permissive
Topics: acp, agent, claude-code, code-generation, codex, opencode, gui, grok-build, pi, ade, deepseek-harness, multi-agent, terminal, vibe-coding, worktrees
Last push: 2026-08-26T16:12:46+00:00

## Health v2 (maintenance only)
Score: 78/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 14
- inputs: {"age_days": 205, "days_push": 7, "days_rel": 7, "gap_med": 0.0, "n_releases_24m": 173}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2990, forks 389 (observed 2026-08-28T04:07:34.492211+00:00)

## What it is
Codeg is a multi-agent AI coding workspace that aggregates sessions from agent CLIs like Claude Code, Codex, and OpenCode into one searchable timeline, and lets a lead agent delegate subtasks to other agent types. It runs as a desktop app, self-hosted server, or Docker container, with native iOS and Android clients for remote monitoring and approvals.

## Use cases
- run claude code and codex in one workspace
- aggregate ai coding agent sessions into one timeline
- let one ai agent delegate tasks to another agent
- manage background coding agent tasks on a to-do board
- control coding agents from my phone
- drive coding agents from telegram or feishu
- run ai coding agents headlessly on a server
- schedule recurring automated agent tasks

## When to choose
- you juggle multiple AI coding agent CLIs and want one unified workspace
- you want agents of different types to collaborate on a single task
- you need to monitor and approve agent work remotely from mobile
- you prefer local-first, self-hosted tooling with no telemetry

## When to avoid
- you only use a single coding agent and are happy with its native terminal UI
- you need a fully stable, GA product — mobile apps are still in testing
- you want a cloud-hosted service rather than something you run yourself

## Facets
- artifact type: application
- maturity: active
- function: agent-framework, chatbot, developer-tools, gui, workflow-automation, scheduling, webhook
- domain: developer-tools, self-hosted, large-language-models
- platform: windows, self-hosted, cross-platform
- tags: multi-agent, ai-coding-agents, claude-code, codex, acp, vibe-coding, session-aggregation, mobile-clients, chat-channels, worktrees, ai-agents, automation, macos, linux, android, ios, docker

## Member repositories
- xintaofei/codeg (main) score 78

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:34.492211+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 2026-08-30T07:31:22.646677+00:00, confidence not recorded.
  - readme: https://github.com/xintaofei/codeg (fetched 2026-08-28T04:07:34.492211+00:00, sha 9eb004d7d026)
  - homepage: https://docs.codeg.app (fetched 2026-08-29T09:46:20.442649+00:00, sha f073ed42e68f)
  - site_page: https://docs.codeg.app/getting-started (fetched 2026-08-29T09:46:20.451787+00:00, sha aec03daf8d89)
  - site_page: https://docs.codeg.app/about (fetched 2026-08-29T09:46:20.453813+00:00, sha a6cb889d9a6d)
  - site_page: https://docs.codeg.app/getting-started/installation (fetched 2026-08-29T09:46:20.455574+00:00, sha cd65d5ebbd34)
- Data as of 2026-08-30T08:39:29.467469+00:00.