BAAI-Agents/Cradle

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.

46
/ 100
Emerging

This framework helps AI models perform complex computer tasks using only screenshots as input and mouse/keyboard actions as output, just like a human would. It enables AI agents to reason, improve, and curate skills across various applications and games, from managing an inbox in Outlook to navigating a virtual world in Red Dead Redemption 2. It's for AI researchers and developers who want to train and evaluate AI agents on general computer control tasks.

2,466 stars. No commits in the last 6 months.

Use this if you are building or researching AI agents that need to interact with any computer software or game using only visual input and standard controls, mimicking human interaction.

Not ideal if you need a simple automation script for a repetitive task or if your application already provides a robust API for direct integration.

AI agent development General computer control Task automation Human-computer interaction Reinforcement learning
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 20 / 25

How are scores calculated?

Stars

2,466

Forks

253

Language

Python

License

MIT

Last pushed

Nov 07, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/agents/BAAI-Agents/Cradle"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.