Yandex Search MCP via browser automation
yandex-search-mcp
An unofficial Playwright-based MCP server that searches Yandex through a browser with authenticated cookies, extracting and ranking page content for an agent
- 4 ★ on the repository
- Last commit
Broad access
Why this level
- Works around Yandex's automated-request protection using cookies and browser stealth mode
- Using a shared or borrowed account risks that account being blocked by Yandex
Install
- Works in
- Services
Manual install
git clone https://github.com/bpawnzz/yandex-search-mcp.git && cd yandex-search-mcp && npm install && npm run buildBuilds the server from source; no ready npm package was found at review time.
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Works around Yandex's automated-request protection using cookies and browser stealth mode
- Using a shared or borrowed account risks that account being blocked by Yandex
This is third-party code. Review the repository files before installing.
What it does
The server offers two tools: yandex_search returns titles, links and snippets from Yandex results, and yandex_search_enhanced additionally fetches the content of found pages, ranks it by relevance and formats it for the model's context with a choice of analysis depth and output format (raw, summarized, synthesized, Q&A-ready). Search runs through a Playwright-driven browser with stealth mode against bot detection and authenticated Yandex account cookies, needed to avoid hitting a CAPTCHA on automated requests. It caches results with per-domain TTLs and guards against result clustering from a single source.
Who it is for. For developers who want Runet search inside an agent and are willing to monitor Yandex ToS compliance themselves.
Good fit when
- You need search specifically in Yandex, not Google, inside an agent workflow
- You want extracted and condensed page content for the model's context, not just a link list
- You have your own Yandex account and are willing to use its cookies to authorize requests
Not a fit when
- Yandex ToS compliance matters to you: the tool explicitly works around bot protection with cookies and browser stealth mode
- You want an official, supported search method: this is an unofficial project with no ties to Yandex
Example request
Search Yandex for the latest news about a new processor launch and summarize three sourcesMCP
- Transport
- stdio
- Authentication
- not required
Environment variables
MAX_PAGES_PER_QUERY- Max pages to fetch content from per query
CACHE_TTL_HOURS- Result cache TTL in hours
Limitations
The author explicitly states this is an unofficial project with no ties to Yandex, Anthropic or Playwright, and that compliance with Yandex's terms of service, rate limits and anti-bot policy is the user's own responsibility. The server works around automated-request protection using authenticated cookies and Playwright stealth techniques, which can get the Yandex account whose cookies are used blocked. It needs its own running Playwright browser and Node.js 18+, with no ready npm package at the time of review, only a from-source build.
How to disable. Remove the yandex-search block from claude_desktop_config.json or your MCP client's configuration.
FAQ
Why does it need account cookies?
Without authenticated cookies Yandex shows a CAPTCHA on automated requests; cookies let the tool avoid it.
Is this an official Yandex server?
No, the README explicitly states the project has no connection to Yandex, Anthropic or Google.
Discussion
No questions or reviews yet.
Related
All Yandex Search toolsYandex search MCP server with offline install
MCP Search Server
A search MCP server on the Yandex Cloud Search API with a monitoring dashboard, Brave fallback and an offline bundle for air-gapped networks
Yandex and Google search MCP server
web-compass
An MCP web-search and content-extraction server via Yandex and Google on top of SerpAPI, geared toward RU/CIS results
Yandex search MCP on .NET with AOT compilation
Yandex Search MCP Server (.NET)
A .NET 10 Yandex search MCP server with native AOT compilation, page-to-Markdown conversion and regex search
Clean search MCP across Yandex, Bing and DuckDuckGo
clean-search-mcp
A spam-filtered search MCP server: Yandex, Bing and DuckDuckGo, a 176K-domain blocklist and result quality scoring
Yandex Search API v2 MCP server
yandex-search-mcp
A self-hosted MCP for Yandex Search API v2: web, image and generative search across all six Yandex indexes
Yandex search MCP server in TypeScript
yandex-search-mcp
A TypeScript MCP server for the Yandex Search API with its own result parser and tests
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Yandex Search MCP via browser automation (this card) | no | not required | 4 | |
| Yandex and Google search MCP server | no | not required | 1 | search through Yandex and Google on SerpAPI |
| Yandex search MCP server in TypeScript | no | API key | 0 | in TypeScript, with its own results parser |
| Yandex search MCP on .NET with AOT compilation | no | API key | 0 | on .NET with a native AOT build |
README in short
The README opens with an explicit disclaimer about its unofficial status and the user's responsibility for Yandex's terms of use. It covers basic and enhanced search tools, Playwright stealth mode, cookie-based authorization, multi-region Yandex support, page content extraction and ranking, smart caching, and setup for Claude Desktop and Cline. It lists steps for a from-source build and getting cookies through a browser extension.