MCP Servers in Yandex AI Studio

Official Yandex demo templates: MCP gateways over cloud functions for YandexART, charts, PDF, arXiv, Kaggle and weather

  • 1 ★ on the repository
  • Last commit
MCP serverOfficial

Extended access

Why this level

  • Deploys cloud functions and an MCP gateway that are billed and need service-account permissions in Yandex Cloud
  • The Webdriver server renders arbitrary web pages via a headless browser, which can touch untrusted content
All reasons and checks
Russian stack

yandex-ai-studio/mcp-servers

Install

Services

Manual install

Text for your agent

Configure the yc CLI for the target cloud and folder, set FOLDER_ID and SERVICE_ACCOUNT_ID in .env, run the chosen server's funcdeploy.ps1 from servers/<name>, then mcpdeploy.ps1 --gateway-name <name> --env-file .env.

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • Deploys cloud functions and an MCP gateway that are billed and need service-account permissions in Yandex Cloud
  • The Webdriver server renders arbitrary web pages via a headless browser, which can touch untrusted content

This is third-party code. Review the repository files before installing.

What it does

A repository from the official Yandex AI Studio organization providing seven MCP server templates, each a Yandex Cloud function behind an MCP gateway: YandexART generates images and returns a link, Grapher renders bar/line/pie charts, MarkdownPDF renders Markdown into PDF, arXiv searches papers and returns details, Kaggle searches public datasets, Weather queries current conditions and a forecast via OpenWeatherMap, and Webdriver fetches rendered web pages via Playwright/Chromium in a private serverless container. These are not locally installable packages but PowerShell scripts for deploying your own MCP gateway in your own Yandex Cloud.

Who it is for. For developers and DevOps engineers who want to deploy their own MCP gateway in Yandex Cloud from Yandex's own template.

Good fit when

  • You have a Yandex Cloud account and want an official example of wrapping a cloud function as an MCP gateway
  • You need YandexART image generation or chart rendering as an MCP tool in your own cloud
  • You need safe web page rendering from a private serverless container

Not a fit when

  • You want a ready server with no deployment of your own: these are deploy templates, not a hosted service
  • You have no access to the Yandex Cloud CLI (yc) or a PowerShell environment for the deploy scripts

Example request

Generate an image via YandexART and build a line chart from this data

MCP

Transport
http
Authentication
API key

Environment variables

FOLDER_ID required
The Yandex Cloud folder to deploy functions and the gateway into.
SERVICE_ACCOUNT_ID required
The service account the gateway runs as.

Limitations

All deploy scripts are PowerShell (.ps1); macOS and Linux need pwsh. You need a configured yc CLI, FOLDER_ID and SERVICE_ACCOUNT_ID, and the cloud functions plus the MCP gateway itself are billed by Yandex Cloud. These are demo materials, not a supported production product, the README calls them simply "useful MCP servers for Yandex AI Studio", with no SLA.

How to disable. Delete the deployed MCP gateway and cloud functions via the Yandex Cloud console or the yc CLI.

FAQ

Is this an official Yandex repository?

Yes, it belongs to the yandex-ai-studio organization on GitHub, described as "demo materials for Yandex AI Studio".

Can I run it without Windows?

The scripts are written in PowerShell; on macOS and Linux you can run them via pwsh.

Discussion

No questions or reviews yet.

Official

Yandex Cloud's official MCP server set: documentation, search, app deployment, serverless functions, containers and a data catalog

MCP serverBroad access49Repository stars
Russian stack

MCP server for basic Yandex Cloud operations: virtual machines, Object Storage buckets and serverless functions from an agent chat

MCP serverExtended access1Repository stars

Yandex Tracker MCP in Yandex AI Studio

Яндекс Трекер в MCP Hub

Official

A Yandex AI Studio MCP Hub template: an agent reads Tracker issue links, creates issues, comments and goals and bulk-changes statuses and queues

MCP serverBroad access
Official

Yandex's official MCP server for web search via Yandex Search API: results with sources and a generative answer

MCP serverExtended access47Repository stars
Russian stack

A 155-tool MCP server for Yandex Direct, Metrika, Wordstat and Webmaster, with self-setup OAuth and read-only and confirm modes

MCP serverBroad access3Repository stars
Russian stack

An unofficial single-file MCP server for Yandex Object Storage: bucket listing, contents and file downloads via the AWS SDK

MCP serverExtended access1Repository stars

README in short

The README describes the overall architecture (a client calls the gateway, which performs a functionCall or httpCall against a backend), lists all seven servers with links to their own READMEs, gives a general step-by-step deploy procedure and a separate flow for the container-based Webdriver server, plus an example .env and the mounts format for attaching Object Storage to a function.

README badge

Report a problem