GroupWeave is the open-source infrastructure for user-owned AI, designed to enable co-creation and co-immersion in generative AI content. Currently developing on NEAR, with plans to integrate Shade agents as customizable, semi-autonomous assistants for multimodal content understanding, moderation, and curation.

To get started with GroupWeave development, you'll need to have Node.js, pnpm, Rust, and Python installed. Once you have the prerequisites, follow the steps below.

1. Clone the repository:

   git clone https://github.com/torus-automations/groupweave.git
   cd groupweave

2. Install dependencies:

   pnpm install

3. Build all packages:

   pnpm build

4. Run an application: To run a specific application, use the pnpm --filter <app-name> dev command. For example, to run the creation app:

   pnpm --filter creation dev

For more detailed instructions on the development environment, commands, and project structure, please refer to our comprehensive developer and agent guide:

📜 AGENTS.md

This project is a monorepo managed with Turborepo and pnpm workspaces.

• apps/: Contains the applications, including Next.js frontends, a React Native mobile app, a Python API, and Rust-based agents.
• packages/: Contains shared packages used by the applications, such as UI components, common types, and configs.

A detailed breakdown of all workspaces can be found in the AGENTS.md file.

We welcome contributions! Please fork the repository, create a branch, and make your changes. Once your feature or fix is ready, please make a pull request for review.

For details on how to work with the shared UI component library, please see our UI Component Guide.

This repository uses a script to generate key documentation files (AGENTS.md, CLAUDE.md, etc.) from a single source of truth: docs.json.

If you make changes that require documentation updates (e.g., adding a new workspace), please:

1. Update the docs.json file.
2. Run the generation script: python generate_docs.py.
3. Commit the changes to docs.json and the generated files.

The following features are currently being developed:

• Betting as a mechanism for content curation/moderation with reward and punishment.
• User-owned AI and personalization with confidential AI models and trusted execution environments.