https://github.com/streamshub/streamshub-mcp
MCP servers for Kubernetes-based streaming platforms — read-only, structured access to infrastructure for LLM-driven diagnostics
https://github.com/streamshub/streamshub-mcp
devops k8s kafka kubernetes mcp mcp-server strimzi
Last synced: 2 months ago
JSON representation
MCP servers for Kubernetes-based streaming platforms — read-only, structured access to infrastructure for LLM-driven diagnostics
- Host: GitHub
- URL: https://github.com/streamshub/streamshub-mcp
- Owner: streamshub
- License: apache-2.0
- Created: 2026-01-28T12:57:13.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2026-04-14T11:00:54.000Z (2 months ago)
- Last Synced: 2026-04-14T12:24:39.693Z (2 months ago)
- Topics: devops, k8s, kafka, kubernetes, mcp, mcp-server, strimzi
- Language: Java
- Homepage:
- Size: 673 KB
- Stars: 0
- Watchers: 0
- Forks: 1
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# StreamsHub MCP
[](https://github.com/streamshub/streamshub-mcp/actions/workflows/build.yml)
[](LICENSE)
> [!WARNING]
> This project is in early alpha and under active development. APIs, tool definitions, and configuration may change without notice.
**Talk to your streaming platform.** StreamsHub MCP gives AI assistants direct access to your Kubernetes-based streaming infrastructure through the [Model Context Protocol](https://modelcontextprotocol.io/). Deploy only the servers you need — each component of your streaming stack gets its own MCP server.
Works with any MCP-compatible client including Claude Code, Claude Desktop, VS Code, Copilot, and more.
## MCP Servers
| Server | Description |
|--------|-------------|
| [Strimzi MCP](strimzi-mcp/) | Manage and troubleshoot Strimzi-managed Kafka clusters on Kubernetes |
Each server has its own README with full documentation — available tools, resource templates, deployment, and configuration.
## Beyond Simple Tools
StreamsHub MCP servers go beyond wrapping APIs as tool calls. They leverage the full MCP specification to create an AI-native operations experience:
- **Prompt templates** — structured diagnostic workflows that guide the LLM step-by-step through complex troubleshooting, telling it exactly which tools to call and in what order
- **Resource templates** — live cluster state exposed as structured context that clients can attach directly to conversations, giving the LLM immediate visibility without requiring tool calls
- **Resource subscriptions** — Kubernetes watches that push real-time notifications when cluster state changes, enabling reactive agents that detect and investigate issues automatically
- **Completions** — dynamic autocomplete for parameters (namespaces, cluster names, topics) powered by live Kubernetes queries
- **Progress and cancellation** — long-running operations like log collection report progress back to the client and support mid-operation cancellation
## Getting Started
Pick an MCP server from the table above and follow its README. Each server can be deployed independently as a standalone application or container.
## Built With
- [Quarkus](https://quarkus.io/) — cloud-native Java framework
- [Fabric8 Kubernetes Client](https://github.com/fabric8io/kubernetes-client) — Kubernetes API access
- [Quarkus MCP Server](https://docs.quarkiverse.io/quarkus-mcp-server/dev/index.html) — MCP protocol implementation
## Development
```bash
mvn compile # Compile all modules (includes checkstyle)
mvn test # Run unit tests
```
## Contributing
Contributions are welcome. To get started:
1. Fork the repository and create a feature branch
2. Run `mvn compile` to verify checkstyle compliance
3. Run `mvn test` to ensure all tests pass
4. Open a pull request
## License
[Apache License 2.0](LICENSE)