Teamcity MCP Server (Nikita Rukavkov)
TeamCity build pipelines and artifact management for AI-powered IDE assistants.
Data last scanned today · Reviewed today
Overview
TeamCity MCP Server allows AI agents to interact directly with JetBrains TeamCity environments. It bridges the gap between CI/CD infrastructure and LLM-driven development by providing structured access to projects, build types, agent statuses, and build artifacts through the Model Context Protocol.
Our verdict
This server receives a security score of 29/100, which indicates a high-risk level. The primary security concern is that it provides full write access to your TeamCity environment without a read-only mode, and it lacks OAuth support, relying entirely on bearer token authentication. Due to the lack of recent maintenance, with the last commit occurring over 235 days ago, and the fact that it is a community-maintained project rather than an official vendor release, it is currently suitable only for non-sensitive, isolated development environments.
- Zero direct dependencies minimize the software supply-chain surface area.
- Fully implements JSON-RPC 2.0 over HTTP and WebSocket protocols.
- Comprehensive build search filters include status, branch, and tags.
- Supports environment-based configuration for Docker and Kubernetes deployment.
- No read-only mode restricts the ability to limit server permissions.
- High security risk due to lack of OAuth and active maintenance.
- Community-maintained without official JetBrains vendor support.
- Security score of 29/100 reflects outdated development activity.
Tools
| Tool | Description | Risk |
|---|---|---|
| trigger_build | Trigger a new build in TeamCity. | medium |
| cancel_build | Cancel a running build. | medium |
| pin_build | Pin or unpin a build to prevent it from being cleaned up. | medium |
| set_build_tag | Add or remove tags from a build. | medium |
| download_artifact | Download build artifacts. | low |
| search_builds | Search for builds with comprehensive filtering options. | low |
| fetch_build_log | Fetch the build log for a specific build with filtering options to handle large logs. | low |
| search_build_configurations | Search for build configurations with comprehensive filtering options including basic filters, parameters, steps, and VCS roots. | low |
| get_current_time | Get the current server date and time to ensure AI models use real current time instead of training data dates. | low |
| get_test_results | Get test results for a specific build with optional filtering by test status. | low |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf | ||||
| ChatGPT |
Frequently asked questions
›What authentication is required to connect to TeamCity?
You must provide a valid TeamCity API token via the TC_TOKEN environment variable.
›How can I secure the communication between the client and the MCP server?
You can configure HMAC authentication by setting the SERVER_SECRET environment variable, which requires the client to include an authorization header.
›Can I deploy this server in a production environment with TLS?
Yes, you can enable TLS by providing the paths to your certificate and private key files via the TLS_CERT and TLS_KEY environment variables.
›Does the server provide caching for API responses?
Yes, it supports response caching, which defaults to 10 seconds but can be adjusted using the CACHE_TTL environment variable.
›Is it possible to configure the server's logging output?
You can customize logs by setting the LOG_LEVEL (e.g., debug, info, warn, error) and the LOG_FORMAT (json or console) environment variables.
Badge
Maintain this server? Add the live badge to your README.