BrowserHive documentation

BrowserHive is a local MCP server that gives AI agents isolated, stealthy Chromium sessions, with vault-backed logins the model never sees, human takeover, a full audit trail and an operator dashboard.

Get started

• Installation: Bun, the package, browserhive init, browserhive doctor
• Quick start: first run, the dashboard, common setups
• Connecting MCP clients: Claude Code, Claude Desktop, Cursor, VS Code, any Streamable HTTP or stdio client

Guides

• Dashboard: a tour of every page
• Configuration: precedence, naming, the config file
• Security model: authentication, bind rules, the vault model, redaction, what is recorded
• Vault: Bitwarden setup, folder policies, bindings, confirmations
• Human takeover: request_attention and the live view
• Notifications: notifications on your phone through Telegram, Discord, ntfy or a webhook, answering from the chat, daily digests and anomaly alerts, reports in the dashboard, screenshots, self-destruct and a public address for links
• Stealth: what it does, what it does not, ceilings, proxies
• Telemetry: OpenTelemetry export and a local Grafana stack
• Command line: every command and exit code
• Programmatic API: createServer
• Upgrading and downgrading: migrations, backups, db restore
• Troubleshooting
• FAQ

Reference

Generated from the source by scripts/gen-docs.ts; always in sync with the code.

• Configuration keys: every key with CLI, environment and file names, type, default
• MCP tools: the 43 tools with parameters, results and errors
• Errors: every error code with cause and resolution
• REST API: every admin API endpoint with scope and authentication
• WebSocket protocol: envelope, topics, commands, screencast frames
• Config file JSON Schema
• Notification message JSON Schema

Contributing

• Architecture: how the code is organised
• Contributing guide