skills.rest
reason-machines
reason-machinesOFFICIAL
Explore all skills by this author →
Website𝕏@reasonmachinesGitHub

gemini-search-mcp-web-search

Provides real-time web search to MCP clients through Google AI Mode via Chrome DevTools Protocol.

7|2|Updated May 16, 2026
Search Tags:#web search#Gemini#MCP server#claude desktop#Chrome DevTools Protocol#api-key-free#google ai mode
One-click install
npx skills add https://github.com/reason-machines/mcp-skills --skill gemini-search-mcp-web-search-reason-machines
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: gemini-search-mcp-web-search
Source: https://github.com/reason-machines/mcp-skills/tree/main/skills/gemini-search-mcp-web-search
Command: npx skills add https://github.com/reason-machines/mcp-skills --skill gemini-search-mcp-web-search-reason-machines

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI agents in Claude Desktop, Cursor, and Windsurf lack access to current web information, and traditional search APIs require paid keys and impose quotas. This Skill connects an MCP server that performs live Google searches through a real Chrome browser, delivering synthesized answers without API keys. ## Core Features & Use Cases - web_search Tool: Executes real-time Google searches and returns synthesized answers grounded in current results. - ask Tool: Handles general questions where Google AI Mode automatically decides whether a web search is needed. - OpenAI-Compatible API: Runs as an HTTP server on port 8080 so non-MCP clients and LangChain integrations can query it. - Use Case: An agent in Claude Desktop needs today's Bitcoin price or the latest AI regulation news; it calls web_search and receives a grounded, up-to-date answer with no API key configuration. ## Quick Start Install the package with pip, then register the gemini-search-mcp command as an MCP server in your client and ask the agent to search the web for a current topic.

Frequently Asked Questions about gemini-search-mcp-web-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add free web search to Claude Desktop without an API key?▼

Install gemini-search-mcp with pip, then add it to claude_desktop_config.json under mcpServers with the command gemini-search-mcp. The server searches Google through a real Chrome browser via CDP, so no search API key is required.

What is the difference between the web_search and ask tools?▼

web_search always performs a live Google search and returns a synthesized answer grounded in results. The ask tool lets Google AI Mode decide whether searching is needed, so pure computations like arithmetic are answered without a web request.

Can I use gemini-search-mcp with Cursor or Windsurf?▼

Yes, add the gemini-search server entry to your MCP settings file such as .cursor/mcp.json with the command gemini-search-mcp. It also runs as an OpenAI-compatible HTTP server on port 8080 for non-MCP clients.

Why does Google show a CAPTCHA when using gemini-search-mcp?▼

Google may flag temporary browser profiles with a /sorry/ CAPTCHA page. Fix it by creating a persistent Chrome profile with GEMINI_SEARCH_USER_DATA_DIR, switching to the undetected-chromedriver backend, or changing BROWSER_CHANNEL to msedge.

Does gemini-search-mcp support proxies and existing Chrome instances?▼

Yes, set GEMINI_SEARCH_PROXY_SERVER to route traffic through a proxy such as socks5. To reuse a running browser, launch Chrome with --remote-debugging-port=9222 and set CDP_URL to http://127.0.0.1:9222.

Related Skills

Explore More Software Engineering Skills→
ihuzaifashoukatihuzaifashoukat
165

x-use

Installs, registers, and configures the x-use MCP server for browser-based X automation.

Community
Intermediate
NousResearchNousResearch
239.1k

pinecone-research

Build agent RAG pipelines and long-term memory using Pinecone vector database.

Official
Intermediate
PytaichukBohdanPytaichukBohdan

reddit-account-creator

Automate Reddit account creation with GoLogin profiles, 2captcha, and PostgreSQL persistence.

Community
Advanced
yangchuanshengyangchuansheng
9

Etsy — Scraping & Data Extraction

Extract Etsy listings and shop data via API and browser scraping.

Community
Advanced