The Yandex Tracker MCP server acts as a bridge between AI assistants and the Yandex Tracker project management platform. In simple terms, it allows tools like Claude, Cursor, or Zed to directly interact with tasks, project queues, and team comments without the user needing to manually copy and paste information. By giving an AI the ability to "see" and "organize" issues, teams can automate routine project management chores and get instant summaries of project progress through natural conversation. Beyond basic task reading, this tool enables comprehensive issue lifecycle management. It provides the AI with deep access to detailed worklogs, attachments, and related issue links, as well as the ability to list and filter through organizational queues. Whether an engineer needs to find a specific bug report or a manager wants a summary of recent activity, the server handles the heavy lifting of API communication, allowing the AI to synthesize data into actionable insights or draft detailed responses based on project history. For developers seeking sophisticated integration, the server supports the full Yandex Tracker Query Language, enabling complex filtering, sorting, and date-based functions. It is built for performance and security, offering optional Redis caching to speed up response times and supporting multiple authentication methods including OAuth 2.0, IAM tokens, and service account credentials. Compatible with both standard Yandex 360 and Yandex Cloud organizations, it can be deployed easily via Docker or the uv package manager, making it a flexible and robust addition to any modern AI-driven development environment.
Prerequisites * Python 3.12 is required. (macOS: brew install python@3.12) * uv (installed globally) or Docker. * Yandex Tracker API Token with appropriate permissions. * Organization ID: Either TRACKER_CLOUD_ORG_ID (for Yandex Cloud) or TRACKER_ORG_ID (for Yandex 360).
One-Click Installation (Claude Desktop) 1. Download the *.dxt file from the latest GitHub Releases. 2. Double-click the file to install it in Claude Desktop. 3. Provide your Yandex Tracker OAuth token when prompted.
2. Configuration
Environment Variables The server requires the following variables: * Authentication (choose one): * TRACKER_TOKEN: Yandex Tracker OAuth token. * TRACKER_IAM_TOKEN: IAM token. * TRACKER_SA_KEY_ID, TRACKER_SA_SERVICE_ACCOUNT_ID, TRACKER_SA_PRIVATE_KEY: Service account credentials. * Organization ID (choose one): * TRACKER_CLOUD_ORG_ID: For Yandex Cloud-managed organizations. * TRACKER_ORG_ID: For Yandex 360 organizations.
3. Available Tools While specific JSON tool definitions aren't listed, the server provides the following capabilities: * Queue Management: List and access all available Yandex Tracker queues with pagination and tag retrieval. * User Management: Retrieve user account info (login, email, license status, organizational data). * Issue Operations: Retrieve detailed issue info, comments, related links, worklogs, and attachments. * Field Management: Access global fields, local fields, statuses, and issue types. * Advanced Search: Support for Yandex Tracker Query Language (complex filtering and sorting).
4. Example Prompts (No specific example prompts were provided in the content)
What you can do with Yandex Tracker
Use Case 1: Automated Daily Standup and Progress Reporting Problem: Developers and project managers often spend significant time manually gathering status updates from various queues and issue boards to prepare for daily standups or weekly reports. Solution: This MCP allows an AI assistant to query Yandex Tracker directly. It can filter issues by assignee, status, and update timestamp across multiple queues to generate a concise summary of accomplishments and blockers. Example: A developer asks Claude: "Summarize everything I moved to 'Done' yesterday in the DEVELOP and QA queues, and list my current 'In Progress' tasks with their deadlines."
Use Case 2: In-IDE Context Gathering for Bug Fixing Problem: When a developer is assigned a bug in an IDE (like Cursor or VS Code), they often have to switch to a browser to read the full description, look at attachments, or check the comment history to understand the context. Solution: By using the Yandex Tracker MCP within a compatible IDE, the AI can fetch the complete issue details, including previous comments and related links, and present them directly alongside the code. Example: A developer highlights a task ID in their code and asks the AI: "Fetch the details for TASK-405. What were the specific reproduction steps mentioned in the latest comments, and are there any linked issues I should be aware of before fixing this?"
Use Case 3: Advanced Issue Auditing and Worklog Management Problem: Ensuring that all tasks are properly logged and that critical issues aren't "rotting" (staying open without updates) is difficult using the standard web interface for large teams. Solution: The MCP leverages the Yandex Tracker Query Language (YQL) support to perform complex searches that identify outliers, such as high-priority bugs without recent comments or tasks missing worklog entries. Example: A team lead asks: "Use a YQL query to find all 'Critical' priority issues in the 'INFRA' queue that haven't been updated in 3 days. Also, check if I have any tasks marked as 'Closed' this week that are missing worklog entries."
Use Case 4: Intelligent Issue Triage and Queue Organization Problem: Incoming tickets in a support or "DevOps" queue can become overwhelming, making it hard to categorize them or find similar historical issues to assist with resolution. Solution: The AI can use the MCP to list all available queues, access global and local fields, and search historical issues to suggest where a new…