Which integrations are available for this server?

Provides search capabilities on TripAdvisor through the platform_search tool, enabling querying for attractions, restaurants, and hotels with site-aware search. Provides search capabilities on Yelp through the platform_search tool, enabling querying for businesses and places with support for site-aware search.

How do I use browser-mcp-demo?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@browser-mcp-demo search XHS for hiking gear and extract first result" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

browser-mcp-demo

by Hanny658

Overview Schema Related Servers Score Discussions

TypeScript

Hybrid

Remote Browser + MCP Tool Gateway (Multi-Site MVP)

This project provides a minimal HITL (human-in-the-loop) remote browser session and an MCP tool gateway for restricted search/extraction. It currently supports XHS and includes real search adapters for Yelp and TripAdvisor.

Requirements

Node.js >= 18
Playwright Chromium (npx playwright install chromium)

Related MCP server: browser-research-mcp

Install

npm install
npx playwright install chromium

Copy the environment template and adjust as needed:

cp .env.example .env

Run

npm run dev

This starts:

HTTP server on http://HOST:PORT
MCP server on stdio (connect with an MCP client)

Docker (single-user + noVNC)

This path is intended for a single user (or MAX_SESSIONS=1). It runs the browser inside Xvfb and streams the desktop via noVNC.

Build:

docker build -t browser-mcp-demo .

Run:

docker run --rm \
  -p 3000:3000 -p 7900:7900 \
  -e HOST=0.0.0.0 \
  -e HEADLESS=false \
  -e MAX_SESSIONS=1 \
  -e VIEW_MODE=novnc \
  -e PUBLIC_BASE_URL=http://YOUR_SERVER_IP:3000 \
  -e NOVNC_URL_TEMPLATE="http://YOUR_SERVER_IP:7900/vnc.html?autoconnect=1&resize=scale&path=websockify" \
  -e PROFILES_DIR=/data/profiles \
  -e AUDIT_LOG_PATH=/data/logs/audit.log \
  -e DELETE_PROFILE=false \
  -v "$PWD/profiles:/data/profiles" \
  -v "$PWD/logs:/data/logs" \
  browser-mcp-demo

Notes:

VIEW_MODE=novnc makes /session/view/:id embed the live browser stream.
Update PUBLIC_BASE_URL and NOVNC_URL_TEMPLATE with your public host or domain.

Call MCP tool create_session -> { sessionId, viewUrl }
Open viewUrl in your browser.
- Default mode: a local Chromium window is opened for login.
- noVNC mode (VIEW_MODE=novnc): the remote browser stream is embedded in the page.
Login on that window (QR/OTP/2FA handled by the user).
Call wait_for_login until status is READY (site-aware when site is provided).

MCP Tools (stdio)

Tools:

create_session
wait_for_login (optional site)
platform_search (site-aware via site param)
xhs_open_and_extract (site-aware via site param)
destroy_session

Example (pseudo):

const session = await client.callTool("create_session", {});
await client.callTool("wait_for_login", { sessionId: session.sessionId, timeoutSec: 120 });
const results = await client.callTool("platform_search", {
  sessionId: session.sessionId,
  query: "camping",
  maxNotes: 10,
  scrollTimes: 0,
  site: "xhs" // xhs | yelp | tripadvisor
});
const detail = await client.callTool("xhs_open_and_extract", {
  sessionId: session.sessionId,
  url: results.notes[0]?.url,
  site: "xhs"
});

Security Boundary

Tools return sanitized structured JSON only.
No cookies, localStorage, sessionStorage, storageState, or userDataDir exposure.
No screenshot tool.
Audit log is written to logs/audit.log with redaction.

Agent HTTP Endpoints

POST /agent/run → start a run and execute until login required or done
POST /agent/continue → continue a run after user login
GET /agent/run/:id → fetch current run state

Example request body:

{
  "query": "camping",
  "maxNotes": 10,
  "scrollTimes": 0,
  "detailCount": 3,
  "detailParallel": 4,
  "site": "xhs"
}

Configuration

Key environment variables:

HOST, PORT, PUBLIC_BASE_URL
UI_DIST_DIR (serve built UI from the same server)
VIEW_MODE (info | novnc)
NOVNC_URL_TEMPLATE (supports {sessionId} placeholder)
OPENAI_API_KEY, OPENAI_MODEL
AGENT_RUN_TTL_MINUTES
MAX_SESSIONS, SESSION_TTL_MINUTES
PROFILES_DIR, DELETE_PROFILE
HEADLESS
XHS_BASE_URL
AUDIT_LOG_PATH

Notes

XHS, Yelp, and TripAdvisor all support search in the current adapter layer. XHS detail extraction is implemented; Yelp/TripAdvisor detail extraction is still stubbed.
The DOM selectors for each site may change. Update src/browser/xhs.ts or src/sites/*.ts if extraction breaks.
This MVP does not implement large-scale crawling or anti-bot bypass.

Available Tools

5 tools

create_sessionC

Create a new browser session and return sessionId + viewUrl.

ParametersJSON Schema

Name	Required	Description	Default
`site`	No

TDQS

C2.7/5.0

Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations provided, so description must bear full weight. Only mentions return values but omits behavioral traits like browser launch, authentication needs, or side effects.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

A single sentence is efficient but too sparse for a session creation tool. Lacks structure to convey essential details while remaining concise.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given one parameter, no annotations, and no output schema, the description fails to explain parameter semantics, return value structure, or when to use. Incomplete for effective agent usage.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters1/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Single parameter 'site' with type string and no description. Schema coverage is 0%, and the description provides no clarification on what 'site' means (e.g., URL, domain, name).

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the action ('Create a new browser session') and the return values ('sessionId + viewUrl'). It distinguishes itself from sibling tools like destroy_session and platform_search.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

No guidance on when to use this tool vs alternatives. Lacks prerequisites or context, such as requiring a valid site parameter or network connectivity.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

destroy_sessionC

Destroy a browser session.

ParametersJSON Schema

Name	Required	Description	Default
`sessionId`	Yes

TDQS

C2.7/5.0

Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries full burden but only implies irreversibility through 'Destroy'. It fails to disclose side effects, required permissions, or what exactly happens to the session.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

At 4 words, it is maximally concise, but the trade-off is insufficient context. The sentence is front-loaded but lacks necessary detail.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Despite low complexity (1 parameter, no output schema), the description omits critical behavioral details and parameter semantics, making it inadequate for reliable agent invocation.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters1/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The schema has 0% description coverage for the single parameter (sessionId). The description adds no meaning, leaving the agent to guess the parameter's purpose or source.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the action ('Destroy') and the resource ('a browser session'), distinctly differentiating it from sibling tools like create_session.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when or when not to use this tool, nor does it mention any prerequisites or alternatives. The agent receives no context for decision-making.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

platform_searchC

Search a site and return a list of notes.

ParametersJSON Schema

Name	Required	Description	Default
`sessionId`	Yes
`query`	Yes
`maxNotes`	No
`scrollTimes`	No
`site`	No

TDQS

C2.4/5.0

Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description does not disclose behavioral traits beyond the basic action. With no annotations, it fails to mention that the tool likely requires an active session, may scroll (implied by scrollTimes), or returns a list. It adds no value over the tool name.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is very short (one sentence), which is concise but at the expense of necessary detail. It is not verbose, but it omits critical information, so a 3 reflects adequate conciseness with insufficient content.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness1/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a tool with five parameters, no output schema, and no annotations, the description is severely incomplete. It does not explain parameter roles, prerequisites (session), or return format, making it insufficient for correct invocation.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters1/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, and the description adds no meaning to the five parameters. The agent cannot infer what sessionId, maxNotes, scrollTimes, or site do from the description alone, which is insufficient.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states a verb ('search') and resource ('site' and 'notes'), making the tool's purpose understandable. However, it does not distinguish from sibling tools like xhs_open_and_extract, which could also retrieve notes. A 4 is appropriate as it is clear but lacks differentiation.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

No guidance is provided on when to use this tool versus alternatives. For example, it does not mention that a session must be created first via create_session, nor that xhs_open_and_extract might be used for single-note extraction. This omission leaves the agent uncertain about context.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

wait_for_loginC

Poll login status for a session (site-specific when provided).

ParametersJSON Schema

Name	Required	Description	Default
`sessionId`	Yes
`timeoutSec`	No
`site`	No

TDQS

C2.8/5.0

Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description must reveal behavioral traits. It only states 'poll login status' without explaining whether the call blocks, times out, or returns immediately. Key behaviors like polling frequency, error handling, and status values are absent.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The single-sentence description is highly concise and contains no filler. Every word contributes to the purpose.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Despite the tool's apparent simplicity, the description omits critical context: return type, blocking nature, error states, and relationship to session lifecycle. With no output schema, the agent lacks information to handle the tool's response properly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters2/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 0% and the description barely explains parameters. It hints at the optional site parameter but fails to describe sessionId (required) or timeoutSec (optional). Consequently, the description adds little value over the raw schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool polls login status for a session and notes site-specific behavior. It is distinct from sibling tools like create_session and destroy_session which handle session lifecycle, though sibling differentiation is not explicit.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives like create_session or platform_search. It does not mention prerequisites (e.g., session must exist) or when polling is appropriate.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

xhs_open_and_extractC

Open a note and extract full detail (site-specific).

ParametersJSON Schema

Name	Required	Description	Default
`sessionId`	Yes
`url`	Yes
`site`	No

TDQS

C2.8/5.0

Behavior1/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations provided, and the description fails to disclose any behavioral traits such as destructive potential, authentication needs, rate limits, or error behavior. For a tool with zero annotation coverage, this is a critical gap.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single sentence with no fluff, but it could include more useful detail without becoming verbose. It is appropriately concise but not maximally informative.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness1/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool has 3 parameters, no output schema, and no annotations, the description is far too brief. It does not explain return values, error handling, or prerequisites for opening a note, leaving agents underinformed.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 0% (no parameter descriptions), but parameter names (sessionId, url, site) are self-explanatory. The description adds 'site-specific' hint but does not elaborate on the optional 'site' parameter or its values.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the action ('Open a note and extract full detail') and the resource ('note'), with 'site-specific' hinting at context. This uniquely identifies the tool's purpose among siblings like create_session, destroy_session, platform_search, and wait_for_login.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

No guidance on when to use this tool vs alternatives, no prerequisites or exclusions. The description only states what it does, leaving the agent to infer usage context.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.

5 tool updatesv0.3.0May 26, 2026
- First observedcreate_session
- First observeddestroy_session
- First observedplatform_search
- First observedwait_for_login
- First observedxhs_open_and_extract

Install Server

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

B3/5.0

Disambiguation5/5

Each tool has a clear and distinct purpose: session creation/destruction, platform searching, login polling, and site-specific extraction. No overlapping functionality.

Naming Consistency2/5

Naming conventions are inconsistent: create_session and destroy_session follow verb_noun, but platform_search is noun_verb, wait_for_login uses a preposition, and xhs_open_and_extract is a long combined verb. No uniform pattern.

Tool Count5/5

Five tools is an appropriate scope for a demo browser MCP server, covering essential workflows without being overwhelming.

Completeness3/5

The set covers a specific workflow (session management, search, login, extraction) for a particular site, but lacks generic browser actions like navigation, clicking, or form filling, which limits its applicability to other contexts.

Maintenance

ActivityInactive

ResponsivenessSyncing

Resources

GitHub Repository

Need Help?

Related Servers

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

luminati-io-brightdata-mcp
ai.smithery
One MCP for the Web. Easily search, crawl, navigate, and extract websites without getting blocked.…
Browser Automation Web Scraping Search
2
Aginx Browser A
net.aginx
Stealth web browser for agents: search, fetch, click, download and type in persistent MCP sessions.
Browser Automation Agent Orchestration
23114
autobrowserapproval-mcp
com.clauxel.autobrowserapproval
A paid remote MCP for AI agent browser approval MCP, built to return verdicts, receipts, usage logs,
Browser Automation Agent Orchestration
2
Tap
io.github.LeonTing1010
Browser MCP for logged-in tasks. Uses your Chrome — credentials stay local. Zero-token replay.
Browser Automation Web Scraping
217

Related MCP Servers

web-scraper-server
Browser Automation Web Scraping
ImYourBoyRoy
A
license
C
quality
C
maintenance
Provides browser automation and web scraping as MCP tools, enabling autonomous URL ingestion, crawling, extraction, and anti-bot handling with interactive browser control.
Updated a month ago (2026-08-03 15:17 UTC)
62
5
MIT
browser-research-mcp
Browser Automation Web Scraping
rsi-ai-platform
A
license
A
quality
B
maintenance
Browser-based research MCP server that drives a real Chromium browser via patched Playwright to access JavaScript-rendered content, dynamic tables, and login-walled dashboards. It provides tools for visiting URLs to extract DOM text and screenshots, and for performing structured data extraction using Anthropic Claude Sonnet.
Updated 3 months ago (2026-06-18 10:51 UTC)
3
Apache 2.0
Real Browser MCP
Browser Automation Web Scraping
qxZap
A
license
Not graded
quality
D
maintenance
Provides a real browser that bypasses bot detection (Cloudflare, Turnstile) for AI agents, enabling navigation, clicking, typing, screenshots, and data collection through MCP tools.
Updated 7 months ago (2026-02-16 09:18 UTC)
104
MIT
AutoBrowser Approval Cloud
Browser Automation Agent Orchestration
clauxel
F
license
Not graded
quality
D
maintenance
Hosted remote MCP for AI agent browser approval. Provides structured tools for page approval workflows, session management, and audit receipts.
Updated 3 months ago (2026-05-25 09:18 UTC)
-

Latest Blog Posts

Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)
By Om-Shree-0709 on July 25, 2026.
mcp
Agent Identity
OAuth 2.1
Your AI Chatbot Just Exposed Your CEO's Salary to an Intern
By Om-Shree-0709 on July 2, 2026.
Agent Identity
MCP Security
OAuth Delegation
Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)
By Om-Shree-0709 on June 30, 2026.
Agentic Ai
Prompt Injection
WebAssembly

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Hanny658/browser-mcp-demo'

If you have feedback or need assistance with the MCP directory API, please join our Discord server

Remote Browser + MCP Tool Gateway (Multi-Site MVP)

Requirements

Install

Run

Docker (single-user + noVNC)

HITL Login Flow

MCP Tools (stdio)

Security Boundary

Agent HTTP Endpoints

Configuration

Notes

Available Tools

Tool Schema Changelog

Maintenance

Resources

Looking for Admin?

Related MCP Connectors

Related MCP Servers

Latest Blog Posts

MCP directory API