Istio MCP‑over‑XDSv3 Server Sample

The Istio MCP-over-XDSv3 server is a reference implementation that demonstrates how to expose Istio configuration data via the Model Context Protocol (MCP) over XDSv3. It bridges the gap between Istio’s control plane and external AI assistants by providing a lightweight, gRPC‑based endpoint that streams configuration resources in real time. This solves the problem of AI tools needing up‑to‑date service mesh metadata—such as virtual services, destination rules, and sidecar configurations—without requiring direct access to Istio’s internal APIs or a full mesh installation.

By running this server inside a Kubernetes cluster, developers can expose the same configuration streams that Istio’s own control plane consumes. The server listens on a standard XDS port (15010) and presents an ‑compatible config source. AI assistants can then subscribe to the MCP stream, receiving incremental updates whenever a resource changes. This guarantees that any AI‑driven automation or policy enforcement remains consistent with the live mesh state, enabling use cases such as dynamic traffic routing suggestions, automated policy compliance checks, and real‑time observability dashboards.

Key capabilities include:

• MCP over XDSv3 compatibility: The server implements the same protocol version used by Istio, ensuring zero‑friction integration with existing mesh deployments.
• Incremental updates: Clients receive only the changes that occur, reducing bandwidth and processing overhead.
• Kubernetes deployment: A single manifest deploys the container image, making it trivial to add to any cluster that already runs Istio.
• ConfigSource integration: By adding the server’s address to Istio’s , the mesh treats it as a native configuration source, allowing seamless discovery of resources.

Typical real‑world scenarios include:

• AI‑powered service mesh management: A Claude assistant can query the MCP stream to suggest optimal routing rules or detect misconfigurations.
• Observability automation: Automated dashboards can pull current mesh topology and metrics from the MCP stream, keeping visualizations in sync with deployments.
• Compliance monitoring: Security tools can listen for policy changes and trigger alerts or remedial actions immediately.

Because the server is a drop‑in replacement for Istio’s own XDS endpoint, it offers unique advantages: minimal operational overhead, full compatibility with existing Istio tooling, and a clear separation of concerns—AI assistants consume data without needing to understand the intricacies of Istio’s control plane. This makes it an attractive component for developers looking to embed intelligent insights directly into their service mesh workflows.