A0 browser use extension#1069
Open
TerminallyLazy wants to merge 13 commits intoagent0ai:developmentfrom
Open
Conversation
Design for a new browser-use plugin that provides step-by-step and autonomous browser tools with CDP screencast viewer in the WebUI. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
12-task implementation plan with complete code for all 14 plugin files, covering session manager, tools, API handlers, CDP proxy, extensions, and WebUI components. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Foundation for the browser_use plugin. Manages a single headed Chromium session per agent context with asyncio.Lock for tool/viewer concurrency, agent.get_data/set_data persistence, and clean shutdown. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Step-by-step browser control tool where each call performs exactly one action and returns the result. Uses the shared SessionManager for browser lifecycle. Includes: open, state, click, type, input, screenshot, scroll, back, keys, select, extract, eval, close. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Wraps browser-use's Agent class with shared SessionManager, configurable vision/max_steps/flash_mode params, secrets masking, cancellation via InterventionException, and asyncio.wait_for timeout. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Browser viewer store manages connection lifecycle, polling for screenshots/status, navigation, and cleanup. Modal HTML provides URL bar with status indicator, canvas for live screenshots, agent busy overlay, and footer with screenshot/close actions. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- SessionManager.ensure_started(): add double-checked locking to prevent concurrent callers from launching duplicate browser instances - Remove __del__ method that unsafely created a new event loop via asyncio.set_event_loop(), which could corrupt the running app loop. Cleanup is already handled by the agent_init extension and explicit close() - Move x-create from outer div to inner template container so $refs.screenCanvas exists when open() fires, and add null guard in store Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
plugins/browser_use/plugin implementing browser automation for Agent Zero's upcoming plugin system (PR feat: Unified Plugin System & Memory Plugin PoC #998)browser_step(13 deterministic actions: open, click, type, scroll, etc.) andbrowser_auto(autonomous browser-use Agent wrapper with configurable max_steps, vision, flash_mode)SessionManagerwith double-checked locking manages a single headed Chromium session per agent contextFiles (14 plugin files + 2 docs)
plugins/browser_use/ ├── api/ │ ├── browser_use_connect.py # Session lifecycle (start/stop/status) │ ├── browser_use_interact.py # HTTP interaction (navigate/screenshot/state) │ └── browser_use_settings.py # Settings CRUD ├── tools/ │ ├── browser_step.py # 13-action deterministic tool │ └── browser_auto.py # Autonomous browser-use Agent wrapper ├── helpers/ │ ├── session_manager.py # Shared browser session lifecycle │ └── cdp_proxy.py # CDP WebSocket proxy with whitelisting ├── extensions/ │ ├── python/agent_init/ │ │ └── _10_browser_cleanup.py # Cleanup on agent init/reset │ └── webui/sidebar-quick-actions-main-start/ │ └── browser-entry.html # Sidebar globe button ├── webui/ │ ├── browser-viewer.html # Viewer modal (canvas + URL bar + overlay) │ ├── browser-viewer-store.js # Alpine store for viewer state │ ├── browser-settings.html # Settings tab component │ └── browser-settings-store.js # Alpine store for settings └── prompts/ └── agent.system.tool.browser_step.mdDependencies
browser-usePython package (already in requirements)