vercel-labs/agent-browser is a browser automation CLI (command-line interface) for AI agents. It drives a web browser programmatically — navigating to pages, clicking on-screen elements and typing into them, reading content and checking it as text, and capturing the screen as screenshots — all from the command line. The latest release in this repository is v0.39.1, and the provided material contains no changelog for this version.
The README describes agent-browser as a fast, native Rust browser automation CLI. It supports page navigation, accessibility tree snapshots, clicking, typing and text verification, and saving screenshots, and it is published under the Apache-2.0 license. The accessibility tree snapshot (a tree that transcribes on-screen elements with their structure intact) is the route through which the screen is read as text, letting the agent find and operate elements. Taken together — the fact that it is a native implementation whose primary language is Rust, the scale of 44,340 GitHub stars, and the fact that the repository was created on January 11, 2026 — it reads as a tool narrowly focused on moving a browser and reading from it.
That said, what can be confirmed extends only as far as the repository description and the feature list spelled out in the README. What changed in v0.39.1, and whether there are speed or stability figures compared with other browser automation tools, are not in the given material. Of the score items contained in the material, only momentum remains, at 7.4 out of 10, while the freshness, technology and quality metrics carry no values. The feature scope and the license are clear, but what this release sets ahead of previous versions cannot be known until the release notes are checked directly.