#952 · Primary category: AI Agents & Automation

opencode-browser

OpenCode plugin to automate Chrome. Inspired by Claude in Chrome.

Project last updated:05/15/26

GitHub Stars

560

Forks

52

Contributors

4

License

MIT

Why we included this project

OpenCode users who want their coding agent to actually touch a live web page rather than just reason about it can get there with this plugin. It talks to a browser over the Chrome DevTools Protocol and hands the agent a compact set of tools: navigate, snapshot the accessibility tree, click, fill, evaluate JavaScript, and take screenshots. That is enough to inspect a page, act on it, and verify the result. The design skips browser extensions and native messaging hosts, and every call takes an explicit browser_url with an optional target_id, so multiple tabs and windows work without hidden state. The same tools run against Chrome, Chromium, Electron, and remote CDP endpoints. For teams building agentic workflows around UI testing, scraping authenticated pages, or driving web apps, the tool list stays small enough to reason about, and setup is just launching a browser with remote debugging enabled.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category