marmot
View on GitHubThe open-source context layer for your AI. Catalog your tables, topics, queues and APIs then expose real metadata to your AI agents.
Marmot is a self-hosted Go data catalog that indexes tables, topics, queues and APIs, then serves that metadata to AI agents through an MCP server. Useful if you want LLM tooling grounded in real, governed data context.
Use Cases
Expose data catalog metadata to AI agents over MCPCatalog database tables, Kafka topics, queues, APIs and dashboardsTrace interactive data lineage and impact analysisFull-text and structured search across data assetsDocument business context, ownership and shared glossariesSelf-host a lightweight data catalog as a single binaryIngest metadata via source pluginsGive LLM agents grounded, certified context about enterprise data
Built With
- Language
- Go
- Frameworks
- Model Context Protocol (modelcontextprotocol/go-sdk) · PostgreSQL · OpenSearch · Kubernetes controller-runtime · Cobra · Viper · OIDC/OAuth2 (fosite) · ORAS · Swagger
Tags
data-catalog · metadata · mcp-server · context-layer · data-lineage · data-discovery · data-governance · data-observability · self-hosted · go · single-binary · plugins · glossary · data-engineering