Kafka Schema Reg MCP

What aywengo/kafka-schema-reg-mcp MCP server does

aywengo/kafka-schema-reg-mcp MCP server connects MCP clients to Kafka Schema Registry operations. Its tool set includes a health check, schema registration, compatibility validation, context creation, single-schema export, and export of all versions for a subject. The README also describes broader capabilities for multi-registry management, schema migration, context-aware administration, and export workflows.

Schema contexts provide logical separation for environments such as development, staging, and production. The server can work with as many as eight registry instances, allowing an agent to operate across environments from one MCP connection. Read-only protection is available through VIEWONLY settings, including separate controls for configured registries.

How it works

The server implements MCP with FastMCP and communicates with clients through JSON-RPC over stdio. Claude Desktop is explicitly supported, and the README states that other MCP clients using the same transport can connect as well. MCP tool calls are translated into requests against the configured Kafka Schema Registry endpoint.

A single-registry deployment uses SCHEMA_REGISTRY_URL. Multi-registry deployments use numbered variables such as SCHEMA_REGISTRY_NAME_1 and SCHEMA_REGISTRY_URL_1, with corresponding per-registry VIEWONLY settings. SLIM_MODE reduces the exposed tool set to an essential subset, which is intended to lower tool overhead for clients and is described as suitable for read-only production operations.

Setup and configuration

The documented deployment options are Docker, a local Python checkout, and Docker Compose. The Docker image is aywengo/kafka-schema-reg-mcp:stable. A basic container run supplies the registry URL, while adding SLIM_MODE=true enables the reduced tool set. Local execution requires Python 3.10 or newer, installation from requirements.txt, and execution of kafka_schema_registry_unified_mcp.py.

The registry username and password variables are optional in the single-registry example. The README also describes OAuth 2.1 authentication with scope-based permissions, but the supplied configuration excerpt does not define the complete OAuth variable set. Prebuilt Claude Desktop configuration examples are provided in the repository for stable Docker and multi-environment setups.

Tools and capabilities

The aywengo/kafka-schema-reg-mcp MCP server exposes capabilities for:

• Checking server health with ping.
• Registering schemas with register_schema.
• Testing compatibility with check_compatibility.
• Creating registry contexts with create_context.
• Exporting one schema with export_schema.
• Exporting every version for a subject with export_subject.
• Managing multiple registry instances and schema contexts.
• Supporting schema migration workflows with backup and verification as described in the README.

Exports are documented as supporting JSON and Avro IDL formats. The project also describes asynchronous operations with progress tracking and cancellation, plus linked navigation in tool responses.

Limitations and notes

The server requires access to a Kafka Schema Registry endpoint; the supplied Docker examples use http://localhost:8081. Registry connectivity, credentials, and permissions must therefore be available to the process. VIEWONLY should be enabled where write protection is required.

The README presents both a 50-plus-tool full mode and a reduced SLIM_MODE, while the listing identifies representative tools rather than a complete inventory. The provided material does not specify every tool name, the full authentication variable set, or detailed request and response schemas. Docker Compose is described as including a Schema Registry for testing, not as a hosted service.