What it does
The server exposes Apache Flink clusters to AI assistants and large language models through the Model Context Protocol, so users can query and manage Flink in conversation. Its tools cover connection setup and status, a cluster overview, a list of jobs with their status, job details by ID, job exceptions, job metrics, TaskManagers with their resources, uploaded JAR files, and sending email notifications such as alerts, status updates or reports.
It suits monitoring cluster and job status, retrieving exceptions and metrics when debugging a job, checking resource use across TaskManagers, and reducing the time spent in the Flink Web UI. An optional connection name lets one server process hold several Flink clusters and target each independently; omitting it uses the "default" connection.
Running it requires a running, accessible Apache Flink cluster, Python 3.8 or higher, and an MCP-compatible client such as Claude Desktop or Continue. The project is released under the MIT License.
Limitations and requirements
The Flink REST API must be enabled, and authentication may be required depending on the Flink setup. The client needs network connectivity to the Flink JobManager. The server provides read-only access to cluster information by default; write operations may require additional configuration and security considerations.
Safety
Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.
- Runs code or changes an assistant's behaviour
Sources
- Official MCP Registry first seen 13 Nov 2026 · last seen 13 Nov 2026 · its page on the source
io.github.Ashfaqbs/apache-flink-mcp-server
MCP server for monitoring Apache Flink clusters via natural language, multi-cluster support.
v1.0.0As shown on registry.modelcontextprotocol.io when we last read it