# 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.

Repository: https://github.com/BAAI-Agents/Cradle
Canonical: https://ross.abutalabs.com/products/cradle
Homepage: https://baai-agents.github.io/Cradle/
Language: Python
License: MIT
License Family: permissive
Topics: ai-agent, ai-agents-framework, computer-control, cradle, gcc, generative-ai, grounding, large-language-models, llm, lmm, multimodality, vision-language-model, vlm, general-computer-control, personoid, ai, foundation-agent
Last push: 2024-11-07T04:58:20+00:00

## Health v2 (maintenance only)
Score: 25/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 65
- inputs: {"age_days": 913, "days_push": 664, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2573, forks 266 (observed 2026-08-28T04:07:01.545704+00:00)

## What it is
Cradle is a Python framework for General Computer Control (GCC), enabling foundation agents to perform complex computer tasks using screenshots as input and keyboard/mouse actions as output. It is a modular LMM-powered framework with modules for self-reflection, task inference, skill curation, action planning, and memory.

## Use cases
- build ai agents that control a computer via screenshots and mouse/keyboard
- automate tasks in games like Red Dead Redemption 2 or Stardew Valley with an llm agent
- create agents that operate desktop software like Chrome or Outlook
- research general computer control with vision-language models
- benchmark foundation agents across diverse virtual environments
- prototype multimodal agents with skill curation and self-reflection

## When to choose
- you want a research framework for agents that interact with GUIs like humans do
- you need a unified screenshot-in, keyboard/mouse-out agent interface across games and software
- you want modular agent components like memory, planning, and skill curation out of the box

## When to avoid
- you need a production-ready RPA tool with guaranteed reliability
- you only need simple browser automation without multimodal reasoning
- you require a lightweight agent without heavy LMM API dependencies

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, computer-vision, llm-inference, workflow-automation
- domain: artificial-intelligence, large-language-models
- platform: python, windows, cross-platform
- tags: general-computer-control, multimodal, vision-language-model, gui-automation, game-agents, screenshot-input, keyboard-mouse-control, ai-agents, automation, desktop

## Member repositories
- BAAI-Agents/Cradle (main) score 25

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:01.545704+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:22:52.120731+00:00, confidence not recorded.
  - readme: https://github.com/BAAI-Agents/Cradle (fetched 2026-08-28T04:07:01.545704+00:00, sha 6e2645a7f644)
  - homepage: https://baai-agents.github.io/Cradle/ (fetched 2026-08-29T10:05:25.196987+00:00, sha a84293764f85)
- Data as of 2026-08-30T08:39:29.467469+00:00.