Yandex Tracker MCP with a report scheduler
Yandex Tracker MCP
A Python Yandex Tracker MCP server with issues, a recurring report scheduler and a SQLite run history
High risk
We rate an entry high when the tool writes to external systems, handles money, production databases or secrets, or runs arbitrary commands. The CLI installs it only with your consent.
Why this level
- Creates, updates and closes issues in the real Tracker when confirmed=true
- Scheduled reports can be sent to an external Telegram bot with real issue data
- Stores the OAuth token and API key in .env
Install
Manual install
py -3.12 -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install -e ".[dev]"Local install on Windows; venv activation commands differ on Linux and macOS.
This is third-party code. Review the repository files before installing.
What it does
The server manages Tracker issues: creating, reading, updating, searching and closing them through available status transitions, since Tracker itself does not allow deleting individual issues. All write calls require confirmed=true, and creation is protected from duplicates via a unique field. A separate tool group creates once, interval or cron schedules that collect a queue summary: total issues, open, overdue, critical and unassigned, saves the report and run to SQLite, and optionally hands the finished Markdown file to an external Telegram bot over a protected endpoint. A three-step composition of search, summarize and save lets the agent take an issue snapshot, compute aggregates and save an artifact for another MCP server.
Who it is for. For teams that need not only Tracker issue operations but also recurring automated queue summaries.
Good fit when
- You need a daily or weekly report on overdue and critical queue issues on a schedule
- You need to safely create and transition issues with protection from accidental duplicates
- You need to compose search, summarize and save an issue snapshot to hand off to another service
Not a fit when
- You need boards and sprints: the server focuses on issues and reports, not board management
- You need direct Telegram sending: the server itself does not send messages, it calls a separate bot service
Example request
Find open issues in the AI queue, summarize the overdue and critical ones and save the reportLimitations
The scheduler and Telegram delivery only work with a separate bot service, which is not part of this repository. Requires an OAuth app with tracker:write scope and a separate MCP_API_KEY to protect its own endpoint. The Docker SQLite storage needs to be kept as a persistent volume.
How to disable. Stop the yandex-tracker-mcp container with docker compose down or remove it from your MCP client config.
MCP
- Transport
- http
- Authentication
- API key
| Environment variables | |
|---|---|
| YANDEX_TRACKER_TOKEN required, secret | OAuth token with tracker:write scope |
| YANDEX_ORG_ID required | Yandex 360 or cloud organization ID, sent as X-Org-ID or X-Cloud-Org-ID |
| YANDEX_DEFAULT_QUEUE required | Default queue for test runs and reports |
| MCP_API_KEY required, secret | Secret protecting the server's public endpoint |
| TELEGRAM_BOT_SERVICE_URL | Address of an external Telegram bot service for report delivery |
Security check
- Creates, updates and closes issues in the real Tracker when confirmed=true
- Scheduled reports can be sent to an external Telegram bot with real issue data
- Stores the OAuth token and API key in .env
README in short
The README lists 12 tools, explains the confirmed and unique safeguards, gives a quick mock-backend run with no risk to real issues, production setup via .env and Docker, a section on getting Tracker access with an OAuth app, the scheduler with Telegram delivery, and a three-step search, summarize, save composition for handing an artifact to another MCP server.
FAQ
Can the server delete an issue?
No, Tracker does not support deleting individual issues; cancel_issue closes or cancels it via a workflow transition instead.
How can I test without risking real issues?
Run the server with TRACKER_BACKEND=mock, so every call goes to a mock backend instead of the real Tracker.
Related
An MCP server with n8n node and template knowledge: the agent picks nodes, validates configs and, with API access, creates workflows in your n8n
Zapier's official MCP plugin: the agent gets actions across thousands of apps through your Zapier account
Awesome Claude Skills by Composio
Awesome Claude Skills
A curated list of Claude skills and plugins, plus Composio's own automation skills for 78 SaaS apps
A plugin and CLI catalog for agents: generates command-line interfaces for GUI apps like GIMP and Blender and installs ready ones via CLI-Hub