Agent-E
An open-source project for automating browser actions through multi-agent task planning and execution.
Description
Agent-E is an open-source project for controlling a browser with natural-language instructions. The repository describes automating actions in the user’s browser.
Its architecture separates planning and execution. The README gives examples such as filling web forms, searching and sorting products, and extracting information from websites.
Running it requires installing the project and configuring a supported language model. Filling PDF forms is listed as unsupported.
Sources: https://github.com/EmergenceAI/Agent-E, https://github.com/EmergenceAI/Agent-E/blob/master/README.md (official sources; reviewed 2026-11-01).