| FMRS | 34 / 100 · D | 66 / 100 · C |
|---|
| Reliability | 4 / 20 | 10 / 20 |
|---|
| Security and permissions | 5 / 20 | 14 / 20 |
|---|
| Maintenance | 10 / 20 | 15 / 20 |
|---|
| Documentation | 8 / 20 | 15 / 20 |
|---|
| Setup experience | 7 / 20 | 12 / 20 |
|---|
| Best for | - Developers who need agents to automate real workflows in logged-in sessions (e.g., personal email, social media, CRM)
- Users who prefer local browser automation with data privacy
- Those who need to run multiple AI agents in parallel and monitor their activity
| - Developers who want AI clients to directly control the browser
- Scenarios requiring DOM-level understanding and low-latency automation (50-200 ms)
- Multi-client support (Claude Code, Cursor, VS Code, etc.)
- Internal tools or automation scripts needing browser interaction
|
|---|
| Not for | - CI/CD testing in stateless, login-free environments (use Playwright or browser-use instead)
- Users who prefer cloud browser services over local execution
- Those who don't need browser automation or only need simple web scraping
| - Headless browser or server-side monitoring
- Bypassing website restrictions or unsupervised automation
- Automation on sensitive pages without human supervision
- Complex scenarios that require visual understanding (e.g., screenshots)
|
|---|
| Required permissions | - Read and import browser logins (from Chrome)
- Access user browser sessions for agent automation
- Store session recordings, screenshots, and history locally under ~/.browserclaw/
- Send anonymous product usage events (agent connect/disconnect, version, OS) – never URLs, page content, prompts, tool results, or screenshots
| - Requires installing the Chrome extension, which requests permissions for active tab, debugger, storage, etc.
- MCP server needs to connect to the extension via WebSocket (default ws://localhost:7225)
- Optional Streamable HTTP mode requires allowing connections to http://127.0.0.1:7226/mcp
- Uploading files requires providing absolute local paths
|
|---|
| Risks and side effects | - Agents use your real login sessions and may perform unintended actions (e.g., sending emails, posting content) – approve with caution
- Malicious prompts could lead to sensitive data exposure or unauthorized actions
- Although data is stored locally, inherent risks of browser automation (e.g., mis-clicks, form errors) remain
- Open-source version is AGPL-3.0 licensed, which may not suit all commercial use cases
| - Browser automation can be unpredictable on complex or sensitive sites; supervision is required
- Credentials and payment info are stored in a vault but require user unlock flows
- API keys are encrypted in Chrome storage, but clients may send page context to model providers
- MCP server is a local process, posing local security risks
- Capability catalog allows most non-denied origins but is gated by multiple checks
|
|---|
| Supported clients | Claude Desktop, Claude Code, Codex, Cursor, VS Code, Zed, OpenCode, Antigravity | Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, Codex |
|---|
| Tools | 0 | 30 |
|---|