feat(opencode): add browser automation - control browsers with 40+ tools #8489
Conversation
## Summary
Adds comprehensive browser automation using Playwright, transforming OpenCode from a local-only coding agent into one that can interact with web applications.
## Browser Control Capabilities
- **Navigation**: navigate, back, forward, refresh, open, close tabs
- **Element Interaction**: click, fill, hover, right-click, double-click, drag-drop
- **Scrolling**: scroll, scrollTo, scrollTop, scrollBottom
- **Form Handling**: check/uncheck, select dropdown, clear, get value
- **Content Extraction**: get text, attributes, CSS, page source, screenshots
- **JavaScript**: execute arbitrary JS in page context
- **Waiting**: wait for elements, wait for URL patterns
- **Storage**: cookies, localStorage, sessionStorage (get/set/delete)
- **Configuration**: viewport, userAgent, geolocation, timezone
- **Assertions**: assert text, visibility, URL patterns
- **Multi-Browser**: Chromium (default), Firefox, Webkit
## Features
- Headed mode by default for visual feedback
- Resource limits (10 pages max, 512MB RAM, 30min idle timeout)
- Retry logic for element staleness
- Permission system requiring user approval for browser actions
## Example Usage
```
browser_navigate({"url": "https://example.com/login"})
browser_fill({"selector": "#email", "value": "user@example.com"})
browser_fill({"selector": "#password", "value": "secret123"})
browser_click({"selector": "button[type=submit]"})
browser_assertURL({"pattern": "/dashboard"})
```
|
Thanks for your contribution! This PR doesn't have a linked issue. All PRs must reference an existing issue. Please:
See CONTRIBUTING.md for details. |
|
The following comment was made by an LLM, it may be inaccurate: Potential Duplicate FoundPR #7302: [FEATURE] Added in-built browser tools using playwright and a parallel playwright node process using spawn for bun-playwright issues Why it's related: This PR also implements browser automation using Playwright with tools for browser control. It appears to address the same core feature of adding browser interaction capabilities to OpenCode, potentially handling similar functionality for controlling browsers with Playwright. You may want to review if PR #7302 is still open/active, and whether PR #8489 is a newer implementation or addresses additional gaps. |
|
im sorry bro 😭😭 |
2 participants
Summary
Adds comprehensive browser automation using Playwright, transforming OpenCode from a local-only coding agent into one that can interact with web applications.
Browser Control Capabilities
Features