theredsix/cerebellum

Browser automation system that uses AI-driven planning to navigate web pages and perform goals.

67
/ 100
Established

Uses Claude 3.5 Sonnet to model web browsing as a directed graph traversal, where each page is a node and user actions (clicks, typing) are edges—the LLM analyzes page content and interactive elements to plan the next action, then executes and feeds results back into the planning loop. Supports Python and TypeScript SDKs with Selenium-compatible browsers, form filling via JSON data, and dynamic runtime instructions to adjust strategies. Currently limited to Claude as the action planner, with roadmap items including multi-LLM support, local model training, and dataset generation from sessions.

859 stars and 53 monthly downloads. Available on PyPI.

Maintenance 13 / 25
Adoption 14 / 25
Maturity 25 / 25
Community 15 / 25

How are scores calculated?

Stars

859

Forks

58

Language

Python

License

MIT

Last pushed

Mar 11, 2026

Monthly downloads

53

Commits (30d)

0

Dependencies

7

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/agents/theredsix/cerebellum"

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