It's a team of sci-fi characters living and working as artificial contributors on Radicle (https://radicle.network). Each character has different skills, a voice and all the characters share a common story. You can configure different models for each one and they work on a pre-configuration basis: you should not worry (or forget) to switch models for different tasks, you can assign a special character with suitable skill set. The expertise is just a way of communication within the team, so you could better understand, which model to choose for a particular task. For example, there is Nico, a junior dev, who could be driven by a fast model like DeepSeek. Or there is Nora, front-end dev, who can use an Anthropic model under the hood.
They are like normal developers, "artificial contributors". You can assign them tasks, they can work on patches, create issues, review patches, and there is an option to configure IRC chat for quick team communication.
Chatting and commenting don't interrupt them from coding, they are independent processes. And they work on a single task at a time - there is a task queue. Tasks can be prioritized. They are sandboxed and you can configure what they can do. All works through the normal Radicle interface, like you would collaborate with normal contributors: patches, issues, labels and mentions.
Right now it uses Claude Code (for Claude) and OpenCode (for other models) under the hood. But I'm considering developing a native harness.
The project is under development and I don't recommend using it for projects in production yet.
Hey everyone! I've open-sourced radicle-team, a project I've been working on for the last couple of months: https://radicle.network/nodes/index.radicle.garden/rad%3Az32... and https://github.com/ales-tsurko/radicle-team
It's a team of sci-fi characters living and working as artificial contributors on Radicle (https://radicle.network). Each character has different skills, a voice and all the characters share a common story. You can configure different models for each one and they work on a pre-configuration basis: you should not worry (or forget) to switch models for different tasks, you can assign a special character with suitable skill set. The expertise is just a way of communication within the team, so you could better understand, which model to choose for a particular task. For example, there is Nico, a junior dev, who could be driven by a fast model like DeepSeek. Or there is Nora, front-end dev, who can use an Anthropic model under the hood.
They are like normal developers, "artificial contributors". You can assign them tasks, they can work on patches, create issues, review patches, and there is an option to configure IRC chat for quick team communication.
Chatting and commenting don't interrupt them from coding, they are independent processes. And they work on a single task at a time - there is a task queue. Tasks can be prioritized. They are sandboxed and you can configure what they can do. All works through the normal Radicle interface, like you would collaborate with normal contributors: patches, issues, labels and mentions.
Right now it uses Claude Code (for Claude) and OpenCode (for other models) under the hood. But I'm considering developing a native harness.
The project is under development and I don't recommend using it for projects in production yet.