---
title: "browse"
description: "Browser automation for AI agents: navigate, extract, screenshot, and interact with real web pages via a single CLI."
canonical: https://agentpluginsdirectory.com/plugins/browse
last-updated: 2026-09-21
---

# browse
Browser automation for AI agents: navigate, extract, screenshot, and interact with real web pages via a single CLI.
- Slug: browse
- Publisher: Browserbase
- Repository: https://github.com/browserbase/browse-plugin
- Manifest: plugin.json
- Version: 0.3.0
- License: MIT
- Category (editorial): research
- Skills: 1 (browse)
- MCP servers: 0
- Stars: 11
- Repository created: 2026-02-11
- Repository last pushed: 2026-09-16
- Publisher type: Organization
- Listing: https://agentpluginsdirectory.com/plugins/browse
- Schema: https://open-plugins.com/schemas/1.0.0/plugin.schema.json

## What browse does, in the publisher's words

Install the browse CLI as a native plugin in Claude Code, Cursor, Codex, Grok, and Gemini CLI.

This repo has no application code. It's a set of static JSON manifests that let each agent marketplace install and SHA-pin the browse plugin, plus the skill that teaches the agent to drive browse from the shell.

See docs/add-a-plugin.md for the full repo layout and how to update the plugin.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/browserbase/browse-plugin/HEAD/README.md

## Skills

- browse: Use the browse CLI for Browserbase browser automation, Browserbase cloud APIs, Browserbase Functions, templates, web fetch/search, diagnostics, and Browse.sh skill discovery/installation. Use when the user asks to navigate pages, inspect browser state, run local or remote browser sessions, manage B…

Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.