github.devotd.
NewCoding agents

spacering-net/codeg v0.38.0 — A workspace for multiple coding agents

spacering-net/codeg is a workspace that handles the conversations, file changes, and sessions of multiple AI coding agents in one place.

Repositoryspacering-net/codeg ↗

spacering-net/codeg is a workspace that handles the conversations, file changes, and sessions of multiple AI coding agents in one place.

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

stars
★ 4K
Forks
519
Language
Rust
License
Apache-2.0
Latest release
v0.38.0
Read original source ↗View on GitHub ↗

spacering-net/codeg is a multi-agent coding workspace that lets you run several AI coding agents in one place and work alongside them. The latest release recorded in the provided repository information is v0.38.0. However, the changelog for this version is not included in the provided materials, so it is not possible to identify what was added or modified.

Codeg brings together agents such as Claude Code, Codex, OpenCode, Pi, and Grok Build, and shows structured conversations and tool cards, files, real-time diffs, and session history together. It supports the bundled agents as well as ACP agents, and sessions can be searched and resumed. Its design focuses on handling the workflows of multiple agents within a single workspace. It is written in Rust and uses the Apache-2.0 license. It can be used as a desktop app, a self-hosted server, and through Docker and on mobile, and connections via Telegram, Lark, and WeChat are also documented.

This material describes the supported features and distribution formats, but it does not contain the specific changes in v0.38.0. Therefore, what differs in this release cannot be determined. No evidence is provided on file-conflict handling between agents, integration stability, or performance comparisons. The scope of handling multiple agents in one place is confirmed, but its robustness in actual operation is difficult to assess from this record alone.