Skip to main content
hc
Submit
mcpSource-backed

bb-browser MCP Server

Browser-control MCP server and CLI that lets AI agents use a real Chrome session, including logged-in browser state, site adapters, DOM snapshots, authenticated fetches, network traces, screenshots, and page actions.

by epiral · submitted by oktofeesh1·added 2026-06-06·
Claude CodeCodexCursorClaude Desktop
HarnessClaude CodeCodexCursorClaude Desktop
Review first — review before installing

Open the source and read safety notes before installing.

Citation facts

Source-backed facts for citing this resource, derived directly from the registry — also available as plain text for AI assistants.

Canonical URL
https://heyclau.de/entry/mcp/bb-browser-mcp-server
Source URLs
https://github.com/epiral/bb-browser/blob/main/README.md, https://github.com/epiral/bb-browser
Brand
bb-browser
Brand domain
github.com
Safety notes
bb-browser can control a real browser session, inspect pages, click elements, fill inputs, evaluate JavaScript, fetch URLs with browser authentication, capture screenshots, and inspect network traffic., Commands may act as the logged-in user on social, finance, job, developer, shopping, and knowledge websites., Keep the daemon bound to trusted local interfaces unless you deliberately configure remote access through a private network., Review community site adapters before using them with authenticated accounts or valuable sessions., Require human approval before posting, submitting forms, changing account settings, applying to jobs, making purchases, or running destructive site actions.
Privacy notes
Tab URLs, titles, DOM snapshots, page content, selected elements, screenshots, network requests, responses, console output, traces, and authenticated fetch results may be exposed to the MCP client and model., Browser login state is used by the page and browser; even if cookies are not extracted directly, commands can act through authenticated sessions., Network request bodies and response data may contain tokens, account data, personal messages, search history, customer information, or internal application data., The upstream privacy doc says communication is local and there is no telemetry or external server, but model providers still receive any data your MCP client sends in prompts or tool results.
Author
epiral
Submitted by
oktofeesh1
Claim status
unclaimed
Last verified
2026-06-06

Decision playbook

Review trust signals before you adopt

Signals are present but mixed. Use the checklist below to confirm the source and operational safety for your environment.

Compare context
Selected

0

Current score

63

Baseline

—

Delta

No baseline selected

No major trust-signal divergence detected in the current selection.

Source and provenance checks

Needs review

Confirm ownership and provenance before trusting install instructions.

  • Source link availableRequired

    Open the canonical repository and verify ownership.

    Done

  • Source provenance statusRequired

    Marked as source-backed.

    Done

  • Metadata reviewed

    No reviewed flag detected in metadata.

    Pending

Safety and privacy checks

Complete

Validate risk disclosures before installation or API wiring.

  • Safety notes presentRequired

    Review the listed safety guidance before running commands.

    Done

  • Privacy notes presentRequired

    Review data handling notes before connecting accounts or secrets.

    Done

  • Trust level risk gateRequired

    Trust level does not block evaluation.

    Done

Package and install checks

Needs review

Check package metadata and artifact integrity signals.

  • Install payload available

    Install or copy payload is available for review.

    Done

  • Package verification flag

    No package verification flag provided.

    Pending

  • Checksum metadata

    No checksum provided for downloaded artifact.

    Pending

Compare-driven decision checks

Needs review

Use compare context to validate trade-offs before adoption.

  • Compare tray has multiple entries

    Add at least one more entry to compare trust differences.

    Pending

  • Baseline comparison available

    No baseline peer selected yet.

    Pending

  • Diverging trust signals identified

    No major trust-signal divergence found.

    Pending

Setup at a glance

CLI install

Copy-ready — paste the snippet to get started.

15 minutes

Install command

Provided

Config snippet

Provided

Copy snippet

Provided

Prerequisites

5 to clear

Platforms

4 listed

Install type

CLI install

Adoption plan

Balanced adoption plan

Current risk score 24/100. Use staged verification before broader rollout.

Risk 24

Pre-adoption checks

Validate source and review signals before any execution.

  • Confirm source provenanceRequired

    Source URL/provenance metadata is present.

    Done

  • Confirm metadata review state

    No review metadata found; increase manual validation.

    Pending

  • Verify install payload

    Install/config payload exists and can be inspected.

    Done

Security checks

Confirm safety, privacy, and package integrity signals.

  • Review safety notesRequired

    Safety notes are present.

    Done

  • Review privacy notesRequired

    Privacy notes are present.

    Done

  • Verify package integrity metadata

    No package verification/checksum metadata.

    Pending

Rollout

Adopt in controlled steps based on the selected plan.

  • Run in isolated sandbox firstRequired

    Use a constrained sandbox and observe behavior across multiple tasks.

    Pending

  • Roll out graduallyRequired

    Roll out to a small cohort before wider usage.

    Pending

  • Set monitoring and fallback

    Define rollback path and monitor errors after adoption.

    Pending

Evidence readiness

Evidence readiness matrix · balanced

Missing required evidence: Metadata review. Risk score 31.

Risk 31

Source provenance

Present

Source repository/provenance is listed.

Required in this preset

Metadata review

Missing

Review metadata is missing.

Required in this preset

Safety notes

Present

Safety notes are present.

Required in this preset

Privacy notes

Present

Privacy notes are present.

Optional in this preset

Package integrity

Missing

Package integrity metadata is missing.

Optional in this preset

Install payload

Present

Install payload is available.

Required in this preset

Required gaps: Metadata review

Decision timeline

Decision timeline · balanced

Blocking gaps: Check metadata review status. Risk 28.

Risk 28

triage

Confirm source provenanceRequired

Source/provenance metadata is available.

Done

triage

Check metadata review statusRequired

Review metadata is missing.

Pending

verify

Review safety notesRequired

Safety notes are available.

Done

verify

Review privacy notes

Privacy notes are available.

Done

verify

Validate package integrity metadata

Package integrity metadata is missing.

Pending

rollout

Verify install payload and commandsRequired

Install payload is available.

Done

Blockers: Check metadata review status

Prerequisite readiness

Prerequisite readiness

5 prerequisites to line up before setup. Have accounts and credentials ready first. Includes a review or approval gate.

0/5 ready
Account & credentials1Install & runtime2Network & hosting1Review & approval115 minutes

Safety & privacy surface

Safety & privacy surface

5 safety and 4 privacy notes across 4 risk areas. Review closely: credentials & tokens, network access, third-party handling.

4 areas
  • SafetyCredentials & tokensbb-browser can control a real browser session, inspect pages, click elements, fill inputs, evaluate JavaScript, fetch URLs with browser authentication, capture screenshots, and inspect network traffic.
  • SafetyExecution & processesCommands may act as the logged-in user on social, finance, job, developer, shopping, and knowledge websites.
  • SafetyNetwork accessKeep the daemon bound to trusted local interfaces unless you deliberately configure remote access through a private network.
  • SafetyCredentials & tokensReview community site adapters before using them with authenticated accounts or valuable sessions.
  • SafetyExecution & processesRequire human approval before posting, submitting forms, changing account settings, applying to jobs, making purchases, or running destructive site actions.
  • PrivacyNetwork accessTab URLs, titles, DOM snapshots, page content, selected elements, screenshots, network requests, responses, console output, traces, and authenticated fetch results may be exposed to the MCP client and model.
  • PrivacyCredentials & tokensBrowser login state is used by the page and browser; even if cookies are not extracted directly, commands can act through authenticated sessions.
  • PrivacyCredentials & tokensNetwork request bodies and response data may contain tokens, account data, personal messages, search history, customer information, or internal application data.
  • PrivacyThird-party handlingThe upstream privacy doc says communication is local and there is no telemetry or external server, but model providers still receive any data your MCP client sends in prompts or tool results.

Safety notes

  • bb-browser can control a real browser session, inspect pages, click elements, fill inputs, evaluate JavaScript, fetch URLs with browser authentication, capture screenshots, and inspect network traffic.
  • Commands may act as the logged-in user on social, finance, job, developer, shopping, and knowledge websites.
  • Keep the daemon bound to trusted local interfaces unless you deliberately configure remote access through a private network.
  • Review community site adapters before using them with authenticated accounts or valuable sessions.
  • Require human approval before posting, submitting forms, changing account settings, applying to jobs, making purchases, or running destructive site actions.

Privacy notes

  • Tab URLs, titles, DOM snapshots, page content, selected elements, screenshots, network requests, responses, console output, traces, and authenticated fetch results may be exposed to the MCP client and model.
  • Browser login state is used by the page and browser; even if cookies are not extracted directly, commands can act through authenticated sessions.
  • Network request bodies and response data may contain tokens, account data, personal messages, search history, customer information, or internal application data.
  • The upstream privacy doc says communication is local and there is no telemetry or external server, but model providers still receive any data your MCP client sends in prompts or tool results.

Prerequisites

  • Node.js 18 or newer.
  • Chrome session, local daemon, and Chrome extension or supported OpenClaw browser path configured according to the upstream README.
  • MCP client such as Claude Code, Codex, Cursor, or another compatible host.
  • Explicit approval to let an agent use the active browser's logged-in state.
  • Site adapter updates reviewed before using community adapters against sensitive accounts.

Schema details

Install type
cli
Troubleshooting
No
Source repository stats
Scope
Source repo
Collection metadata
Estimated setup
15 minutes
Difficulty
advanced
Full copyable content
npx -y bb-browser --mcp

About this resource

Content

bb-browser is a browser-control CLI and MCP server for AI agents. It connects agents to a real Chrome session through a local daemon and browser extension, so the agent can inspect pages, use site adapters, capture screenshots, evaluate page JavaScript, inspect network traffic, and call authenticated browser fetches with the active user's login state.

The project positions the browser itself as the API. That makes it powerful for sites without public APIs, but also means the agent can interact with real accounts and logged-in pages.

Source Review

These sources were reviewed on 2026-06-06. Prefer the live repository, README, privacy document, package metadata, npm registry metadata, and community adapter repository for current MCP configuration, daemon behavior, site commands, privacy claims, and adapter coverage.

Features

  • Run an MCP server with npx -y bb-browser --mcp.
  • Control a real Chrome session through a local daemon and extension.
  • Use site adapters for search, social, news, developer, video, finance, jobs, knowledge, shopping, and tool workflows.
  • Inspect DOM snapshots and accessibility trees.
  • Click elements, fill inputs, evaluate JavaScript, and open pages.
  • Capture screenshots and network request data.
  • Fetch URLs using the browser's authenticated session.
  • Use OpenClaw mode for supported browser workflows without the Chrome extension.
  • Update and recommend community site adapters.

Installation

Configure bb-browser in an MCP client:

{
  "mcpServers": {
    "bb-browser": {
      "command": "npx",
      "args": ["-y", "bb-browser", "--mcp"]
    }
  }
}

Follow the upstream README for the daemon, Chrome extension, OpenClaw path, and site adapter setup. Start with read-only commands such as page inspection, search, or summaries before enabling page actions.

Use Cases

  • Let an agent inspect a logged-in page without separate API credentials.
  • Search or summarize data from websites that do not provide useful public APIs.
  • Capture page structure, screenshots, console output, and network context for debugging.
  • Use community adapters for social, developer, finance, video, job, and knowledge sites.
  • Build new site adapters from observed browser network and DOM behavior.

Safety and Privacy

bb-browser can act through a real browser session. Treat every command as if it could use the privileges of the logged-in human. Keep the daemon local, use non-critical accounts for experiments, and require approval before any command that posts, submits, purchases, deletes, changes settings, or sends messages.

The upstream privacy document says bb-browser communication is local and does not send telemetry or data to an external service. That does not make tool outputs automatically safe for models. DOM text, screenshots, network bodies, authenticated fetch results, and adapter output can still be sent to the MCP client and model provider.

Duplicate Check

No epiral/bb-browser entry, bb-browser MCP entry, or matching source URL was found in content/mcp.

#browser#chrome#automation#web#local-first

Source citations

Add this badge to your README

Show that bb-browser MCP Server is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.

Listed on HeyClaude
[![Listed on HeyClaude](https://heyclau.de/badge/mcp/bb-browser-mcp-server.svg)](https://heyclau.de/entry/mcp/bb-browser-mcp-server)

How it compares

bb-browser MCP Server side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.

Field
bb-browser MCP Server

Browser-control MCP server and CLI that lets AI agents use a real Chrome session, including logged-in browser state, site adapters, DOM snapshots, authenticated fetches, network traces, screenshots, and page actions.

Open dossier
Firefox DevTools MCP

Mozilla-maintained MCP server for automating Firefox through WebDriver BiDi, with tools for page navigation, snapshots, UID-based input, screenshots, network requests, console messages, dialogs, history, viewport changes, optional JavaScript evaluation, privileged Firefox contexts, preferences, and WebExtension.

Open dossier
AgentQL logoAgentQL MCP Server

AgentQL MCP server for extracting structured JSON from public webpages using a URL and natural-language extraction prompt.

Open dossier
Apify logoApify MCP Server

Official Apify MCP server for connecting Claude to Apify Actors, Actor runs, datasets, key-value stores, and Apify documentation through hosted Streamable HTTP or local stdio transports.

Open dossier
Next steps
Open dossierAPI JSONOpen LLMMCP feedOpen sourceNewsletterClaim listing
Open dossierAPI JSONOpen LLMMCP feedOpen sourceNewsletterClaim listing
Open dossierAPI JSONOpen LLMMCP feedOpen sourceNewsletterClaim listing
Open dossierAPI JSONOpen LLMMCP feedOpen sourceNewsletterClaim listing
Trust
Review statusNot reviewedNot reviewedNot reviewedNot reviewed
Package trustPackage not verifiedPackage not verifiedPackage not verifiedPackage not verified
Source provenanceSource-backedSource-backedSource-backedSource-backed
Submitteroktofeesh1oktofeesh1oktofeesh1oktofeesh1
Install riskReview firstReview firstReview firstReview first
Notes Safety ✓ Privacy ✓ Safety ✓ Privacy ✓ Safety ✓ Privacy ✓ Safety ✓ Privacy ✓
Brand——AgentQL logoAgentQLApify logoApify
Categorymcpmcpmcpmcp
SourceSource-backedSource-backedSource-backedSource-backed
AuthorepiralMozillaAgentQLApify
Added2026-06-062026-06-062026-06-062026-06-06
Platforms
Claude CodeCodexCursorClaude Desktop
Claude CodeClaude Desktop
Claude CodeClaude Desktop
Claude CodeClaude Desktop
Harness
Claude CodeCodexCursorClaude Desktop
Claude CodeClaude Desktop
Claude CodeClaude Desktop
Claude CodeClaude Desktop
Source repo————
Safety notes✓bb-browser can control a real browser session, inspect pages, click elements, fill inputs, evaluate JavaScript, fetch URLs with browser authentication, capture screenshots, and inspect network traffic. Commands may act as the logged-in user on social, finance, job, developer, shopping, and knowledge websites. Keep the daemon bound to trusted local interfaces unless you deliberately configure remote access through a private network. Review community site adapters before using them with authenticated accounts or valuable sessions. Require human approval before posting, submitting forms, changing account settings, applying to jobs, making purchases, or running destructive site actions.✓Browser pages can contain prompt-injection text in visible content, hidden HTML, ARIA labels, console output, network responses, or page metadata. The server can navigate pages, click, hover, fill forms, drag, upload files, accept dialogs, dismiss dialogs, go through history, change viewport size, close tabs, and restart Firefox. Enabling evaluate_script allows arbitrary JavaScript in page context; compromised instructions could read page data, modify the DOM, or interact with accessible browser APIs. Enabling privileged context tools with the required Firefox system-access environment variable can expose privileged Firefox APIs and may extend beyond web-content sandbox boundaries. Connect-existing mode can attach to a real browsing session with cookies, logins, active tabs, history, and saved state, so avoid it unless the profile is dedicated to automation. Accepting insecure certificates weakens TLS validation and can hide man-in-the-middle or misconfiguration signals.✓AgentQL MCP exposes one tool, `extract-web-data`, that sends a target URL and natural-language extraction prompt to the AgentQL API. The tool is intended for public webpages; do not use it to bypass access controls, scrape private pages, evade paywalls, or extract data where automated collection is prohibited. Web extraction can still trigger target-site rate limits, legal restrictions, robots guidance, or terms-of-service concerns. The source implementation uses AgentQL's query-data endpoint with fast mode, no screenshot capture, no scroll-to-bottom behavior, and no local browser cookies. Treat extracted output as untrusted web data that may include errors, stale content, ads, tracking text, or prompt-injection attempts.✓Apify MCP can run Apify Actors that browse, scrape, crawl, automate websites, call third-party services, and produce billable Actor runs. The hosted server supports OAuth and bearer-token authentication; local stdio uses APIFY_TOKEN in the MCP client environment. Actor tools can access external websites and may trigger anti-abuse systems, platform-policy violations, rate limits, or paid usage. The server exposes run, dataset, key-value store, and abort tools when Actor workflows are enabled. Agentic payments can fund Actor runs through x402 USDC on Base or Skyfire PAY tokens, so spending limits and human approval should be enforced outside the model. Telemetry is enabled by default for tool calls, and the stdio transport uses Sentry for error tracking unless telemetry is opted out.
Privacy notes✓Tab URLs, titles, DOM snapshots, page content, selected elements, screenshots, network requests, responses, console output, traces, and authenticated fetch results may be exposed to the MCP client and model. Browser login state is used by the page and browser; even if cookies are not extracted directly, commands can act through authenticated sessions. Network request bodies and response data may contain tokens, account data, personal messages, search history, customer information, or internal application data. The upstream privacy doc says communication is local and there is no telemetry or external server, but model providers still receive any data your MCP client sends in prompts or tool results.✓Page content, DOM snapshots, accessibility labels, console messages, network requests, headers, screenshots, uploaded file paths, extension names, Firefox prefs, and browser logs can be sent to the MCP client and model. A regular Firefox profile can expose cookies, saved sessions, browsing history, account data, extensions, and private tabs to browser automation. Screenshot save paths, uploaded files, downloaded artifacts, and model transcripts can reveal local project names, test data, credentials shown on pages, or customer information. Use a separate profile, minimize enabled capabilities, avoid sensitive sites, and review provider/browser data handling before automating logged-in sessions.✓Target URLs, extraction prompts, API key-authenticated requests, and extracted structured data are sent to AgentQL's API. Extracted data can include personal data, copyrighted content, customer information, job postings, prices, social content, or other third-party material. AGENTQL_API_KEY should stay out of prompts, issues, logs, screenshots, and committed configuration files. Claude transcripts and downstream reports may retain extracted data, so avoid collecting information that is not approved for the model session.✓Actor inputs and requests are sent to the Apify API for execution, and returned datasets or key-value store records can contain scraped personal data, customer records, credentials, contact details, private URLs, or regulated information. APIFY_TOKEN, OAuth tokens, Skyfire API keys, wallet private keys, payment IDs, dataset IDs, key-value store keys, and Actor run IDs should not be pasted into prompts, issues, logs, screenshots, or committed files. Search, maps, social, ecommerce, and browser Actors can collect third-party content that may be subject to site terms, privacy law, copyright, or data-retention obligations. Telemetry and Sentry error tracking should be reviewed before using the local stdio server with sensitive prompts, Actor inputs, or operational data.
Prerequisites
  • Node.js 18 or newer.
  • Chrome session, local daemon, and Chrome extension or supported OpenClaw browser path configured according to the upstream README.
  • MCP client such as Claude Code, Codex, Cursor, or another compatible host.
  • Explicit approval to let an agent use the active browser's logged-in state.
  • Node.js 20.19.0 or newer.
  • Local Firefox 100 or newer, or an explicit Firefox binary path passed through the documented CLI option.
  • A dedicated Firefox automation profile with no personal cookies, saved passwords, or regular browsing sessions.
  • Optional adb and Firefox for Android setup when using Android automation mode.
  • Node.js and npx available to the MCP client runtime.
  • AgentQL API key from the AgentQL developer portal.
  • Approved list of public webpages or domains Claude may query.
  • Review of target site terms, robots guidance, rate limits, and data-use rules before extraction.
  • Node.js 22 or newer available in the MCP client runtime for local stdio usage.
  • Apify account, OAuth session, API token, x402 wallet, or Skyfire payment setup depending on the chosen transport and payment mode.
  • Explicit tool selection for production use, because the default tool configuration may change in future versions.
  • Team approval policy for Actor runs, paid scraping workflows, agentic payments, dataset access, and aborting active runs.
Install
npx -y bb-browser --mcp
npx -y firefox-devtools-mcp@latest
npx -y agentql-mcp
npx @apify/actors-mcp-server --tools actors,docs,apify/rag-web-browser
Config
{
  "mcpServers": {
    "bb-browser": {
      "command": "npx",
      "args": ["-y", "bb-browser", "--mcp"]
    }
  }
}
{
  "mcpServers": {
    "firefox-devtools": {
      "command": "npx",
      "args": [
        "-y",
        "firefox-devtools-mcp@latest",
        "--headless",
        "--viewport",
        "1280x720"
      ],
      "env": {
        "START_URL": "about:home",
        "FIREFOX_HEADLESS": "true"
      }
    }
  }
}
{
  "mcpServers": {
    "agentql": {
      "command": "npx",
      "args": ["-y", "agentql-mcp"],
      "env": {
        "AGENTQL_API_KEY": "<your-agentql-api-key>"
      }
    }
  }
}
{
  "mcpServers": {
    "apify": {
      "command": "npx",
      "args": [
        "-y",
        "@apify/actors-mcp-server",
        "--tools",
        "actors,docs,apify/rag-web-browser"
      ],
      "env": {
        "APIFY_TOKEN": "<your-apify-token>"
      }
    }
  }
}
Citations
ClaimUnclaimedUnclaimedUnclaimedUnclaimed
Open 4 picks in the interactive comparison tool

Related guides

Source-backed guides for putting this to work.

Chrome Integration for Web App Debugging With Claude Code

Debug web apps with Claude in Chrome and Claude Code together.

Added 1mo ago
guidesReview firstSource-backedReview first
Safety ✓ Privacy ✓by kiannidev

Auto Mode Hard-Deny Policies For Safe Automation

Set autoMode.hard_deny rules to block risky actions in auto mode.

Added 1mo ago
guidesReview firstSource-backedReview first
Safety ✓ Privacy ✓by kiannidev

Claude Process Automation

Automate business and engineering processes with headless Claude Code, GitHub Actions, and scheduled routines.

Added 8mo ago
guidesReview firstSource-backedReview first
Safety ✓ Privacy ✓by JSONbored

Signals

Loading live community signals…

More like this, weekly

A short, calm digest of reviewed Claude resources. Unsubscribe any time.