What problem does it solve?
Many web tasks require repetitive manual interaction such as navigating pages, filling forms, clicking controls, capturing screenshots, and extracting on-page data; agent-browser streamlines these tasks into deterministic CLI-driven actions to speed testing, demos, and data collection.
Core Features & Use Cases
- Navigation & Snapshotting: Navigate to pages and produce accessibility-style snapshots that return stable element refs for scripted interactions.
- Interactions & Form Automation: Click, fill, type, select, upload, drag-and-drop, and control keyboard/mouse for end-to-end form submission and UI flows.
- Observation & Extraction: Capture screenshots, PDFs, video recordings, and extract text, attributes, and bounding boxes for QA evidence and scraping.
- Sessions & Network Control: Save/load session state, manage multiple parallel sessions, intercept and mock network requests, and run in headed or headless modes.
- Use Case: Use agent-browser to automate login, submit a form, verify dashboard content, record the flow as a demo video, and export element data for reporting.
Quick Start
Use agent-browser to open a webpage, snapshot interactive elements, fill the required fields, submit the form, and save a screenshot.
