Tracker report Telegram delivery MCP
Telegram Delivery MCP
A companion MCP server that takes a finished Yandex Tracker report and sends it to Telegram through an existing bot service
- 1 ★ on the repository
- Last commit
Broad access
Why this level
- The server triggers sending documents to Telegram on behalf of an existing bot
- It requires matching API keys for two external services; a leak of these would allow delivering arbitrary artifacts
Install
- Works in
- Services
Manual install
docker compose up -d --buildRun inside a shared stack, with Tracker MCP and Telegram bot service addresses set in .env.
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The server triggers sending documents to Telegram on behalf of an existing bot
- It requires matching API keys for two external services; a leak of these would allow delivering arbitrary artifacts
This is third-party code. Review the repository files before installing.
What it does
The server is built as a second, independent MCP for a training exercise on orchestrating multiple servers. It does not search Tracker issues itself: the agent first calls search, summarize and save-report on a separate YandexTrackerMCP server and gets an artifact_id, then calls send_tracker_artifact here. This server pulls the markdown document from Tracker MCP via an internal MCP client, hands it to an existing Telegram bot service, records the delivery in SQLite, and returns a delivery_id. Status can be checked with get_delivery_status, and recent attempts with list_recent_deliveries. The Telegram bot's own token is not needed here: the server talks to a protected internal endpoint of an already-deployed bot service.
Who it is for. For those already building or studying orchestration of several MCP servers around Yandex Tracker and existing Telegram bot infrastructure.
Good fit when
- You already run a separate YandexTrackerMCP server producing artifact_id
- You already have a working Telegram bot service with a protected internal endpoint
- You are studying the pattern of orchestrating several MCP servers in one scenario
Not a fit when
- You are looking for a standalone server that both reads Tracker issues and sends to Telegram itself: both pieces here are external and not part of this repository
- You have no ready Telegram bot service with the required internal API: without it send_tracker_artifact will not work
Example request
Find open critical issues in the AI queue, prepare a report and send it to TelegramMCP
- Transport
- http
- Authentication
- API key
Environment variables
TRACKER_MCP_URLrequired- Address of the external YandexTrackerMCP server
TRACKER_MCP_API_KEYrequired, secret- Key matching MCP_API_KEY on the Tracker MCP side
TELEGRAM_BOT_SERVICE_URLrequired- Address of the existing Telegram bot service
TELEGRAM_BOT_SERVICE_API_KEYrequired, secret- Key matching BOT_SERVICE_API_KEY on the bot's side
Limitations
The server is useless on its own: it is designed to pair with an external YandexTrackerMCP (address set via TRACKER_MCP_URL) and an external Telegram bot service (TELEGRAM_BOT_SERVICE_URL), neither of which is part of this repository or documented as publicly available. It is built as a training project for a specific orchestration exercise, not as a general-purpose tool. The two services' API keys must match on both sides, or delivery will fail.
How to disable. Stop the telegram-delivery-mcp container or server process and remove it from the orchestration.
FAQ
Does this server need a Telegram bot token?
No, it calls a protected internal endpoint of an already-deployed bot service, not the Telegram Bot API directly.
Can the server find Tracker issues itself?
No, search and summarization happen on a separate YandexTrackerMCP server; this one only delivers the finished artifact.
Discussion
No questions or reviews yet.
Related
All Yandex Tracker toolsYandex 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
A Yandex Tracker MCP server: queues, issues, comments, status transitions, sprints and search in Tracker query language
yandex-tracker-mcp: Tracker issues, checklists and time tracking
yandex-tracker-mcp
An independent Python MCP server for the Yandex Tracker REST API v3: issues, transitions, checklists, worklogs and time reports
Yandex Tracker MCP without PII
Yandex Tracker MCP
A Yandex Tracker MCP server that deliberately withholds assignees and authors, returning only a whitelist of issue, comment and link fields
Yandex Tracker MCP in Yandex AI Studio
Яндекс Трекер в MCP Hub
A Yandex AI Studio MCP Hub template: an agent reads Tracker issue links, creates issues, comments and goals and bulk-changes statuses and queues
Yandex Tracker MCP: 23 tools for issues and files
yandex_tracker_mcp
A compact Python Yandex Tracker server: 23 tools, attachments and a one-variable read-only mode
README in short
The README describes the server as an independent second MCP for the 'Day 20. Orchestration MCP' training assignment, lists its four tools, explains that SQLite stores deliveries in data/deliveries.db, gives local run instructions via Python and Docker Compose, the address and key variables for the two external services, and an end-to-end verification scenario via a separate e2e script.