#875 · Primary category: AI Agents & Automation

Cradle

ai ai-agent ai-agents-framework computer-control cradle foundation-agent gcc general-computer-control generative-ai grounding large-language-models llm lmm multimodality personoid vision-language-model vlm

The Cradle framework is a first attempt at General Computer Control (GCC). Cradle supports agents to ace any computer task by enabling strong reasoning abilities, self-improvment, and skill curation, in a standardized general environment with minimal requirements.

Project last updated:11/07/24

GitHub Stars

2.6K

Forks

266

Contributors

7

License

MIT

Why we included this project

Cradle is a research framework for what the BAAI team calls General Computer Control: the agent takes a screenshot as its only input and acts by sending keyboard and mouse events, so it can drive software that exposes no special API. The repository ships working environments for a varied set of targets, including the games Red Dead Redemption 2, Stardew Valley, and Cities: Skylines, plus everyday desktop tools like Chrome, Outlook, and CapCut. The interesting part is the skill system: you define small atomic actions, say moving a character or clicking a button, and the agent learns to chain them into larger skills it records and reuses later. If you want to prototype computer-use agents, run long-horizon autonomy experiments, or build GUI automation on top of a vision-language model rather than hard-coded browser hooks, this is a practical starting point.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category