Yandex Wiki Search MCP Server
Yandex Wiki data access including full-text search, page management, and table operations.
Data last scanned today · Reviewed today
Overview
Yandex Wiki Search MCP provides a suite of 33 tools for interacting with Yandex Wiki environments. It supports comprehensive full-text searches, enabling users to query wiki pages, comments, and attachments directly from their AI assistant. Users can execute CRUD operations on wiki pages and manage complex dynamic table grids with specialized tools for row, column, and cell manipulation. The server features a server-side read-only mode, which can be enforced by setting the environment variable WIKI_READ_ONLY=true. It uses typed JSON schemas for all inputs and outputs to ensure data consistency and supports both stdio for local desktop clients and streamable HTTP via Docker for team deployments.
Our verdict
This server carries a high-risk rating due to a security score of 35/100, primarily because it is a community-maintained project rather than an official Yandex release. Additionally, critical security components like authentication methods have not yet been independently verified. Potential users should be aware of these limitations and perform their own due diligence, especially regarding the use of OAuth tokens in professional or sensitive wiki environments.
- Supports 33 distinct tools for wiki and grid management.
- Enforces read-only mode via server-side environment configuration.
- Includes full input and output typed JSON schemas.
- Zero direct dependencies minimize the supply-chain footprint.
- Security score is low at 35/100 with high-risk status.
- Not maintained or endorsed by the official Yandex organization.
- Authentication and read-only implementation are not yet reviewed.
Tools
| Tool | Description | Risk |
|---|---|---|
| page_search | Full-text search across the entire Wiki (pages and files), ranked results with a text excerpt each; server-side filters, and cursor paging. | low |
| page_get | Get a page by page_id or slug (accepts full Wiki URLs too). | low |
| page_get_descendants | Traverse a page subtree — one flat list of {id, slug} from all nesting levels. | low |
| page_get_comments | List page comments (fetch_all supported). | low |
| page_get_resources | List page resources (attachments + grids) with server-side title search (fetch_all supported). | low |
| page_get_attachments | List page attachments (fetch_all supported). | low |
| page_read_attachment | Read an attachment's content straight into the conversation. | low |
| page_get_grids | List grids attached to a page (fetch_all supported). | low |
| grid_get | Get a grid by grid_id with row/column/revision filters. | low |
| user_get_current | Who am I — username and home_cluster. | low |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| ChatGPT | ||||
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf |
Frequently asked questions
›What configuration is required to run the server?
You must provide a Yandex OAuth token and your organization ID via environment variables named WIKI_TOKEN and WIKI_ORG_ID.
›How can I prevent the AI agent from making unwanted changes to my wiki?
You can set the WIKI_READ_ONLY environment variable to true, which causes the server to not register any write tools.
›Is it possible to run this server without a local Python installation?
Yes, you can run the server using Docker by pulling the image from ghcr.io/dlbolshov/yandex-wiki-search-mcp.
›Which MCP SDK version does the server use?
The server uses the MCP Python SDK v2, which maintains backward compatibility with previous protocol revisions.
›Does this server have official support from Yandex?
No, this is an unofficial project and is not affiliated with or endorsed by Yandex.
Alternatives
Changelog
Badge
Maintain this server? Add the live badge to your README.