A collection of example scripts and custom connectors demonstrating OpenMetadata SDK usage for ingesting and managing metadata assets
This is NOT an MCP server. The repository contains OpenMetadata ingestion connectors and demo scripts, not MCP tool definitions. No MCP protocol implementation detected. Tools are Python functions/classes (CustomDatabaseSource, CustomDashboardSource, etc.) designed for OpenMetadata's ingestion framework, not MCP. Input schemas shown are inferred from docstrings/type hints, not formal MCP schemas. Descriptions are present but generic. No error handling, no pagination, no structured output schemas. This appears to be a data ingestion toolkit, not an agent-accessible MCP server.
Yields two charts, a data model and one dashboard referencing both
Yields a tagged catalog of databases, schemas, tables, a stored procedure and lineage
Yields one directory with a file, and one spreadsheet with a worksheet
Yields two topics with a message schema and sample data
Create representative analytics assets with public OpenMetadata facades. Demonstrates creating dashboards, charts, data models, ML models, pipelines, queries, and search indexes
Register API Collections and Endpoints under two existing API services and add lineage edges between endpoints across services. Reads from declarative YAML files for CI/CD pipeline integration
Not an MCP server. Repository is an OpenMetadata ingestion toolkit, not an MCP implementation. No MCP protocol transport, no tool registration via MCP, no MCP schema definitions.
Input schemas are inferred from type hints (e.g., 'config': WorkflowSource, 'metadata': OpenMetadata), not formal JSON Schema with explicit type/description pairs. Cannot verify schema completeness.
Tool names do not follow verb_noun convention. 'CustomDatabaseSource', 'CustomDashboardSource' are noun-heavy and do not clearly signal action. Should be 'ingest_database_catalog', 'ingest_dashboard_metadata', etc.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 39 | 2026-07-28+ | v2 |
Create a small database catalog with the public OpenMetadata SDK facades. Demonstrates creating database services, databases, schemas, tables, and stored procedures with ownership markers
No output schemas documented. Tools yield/return data but no structured schema is declared for LLM consumption. Agents cannot plan downstream operations.
No error handling or recovery guidance. Functions raise exceptions (e.g., ValueError for missing env vars) but do not provide actionable error messages or retry strategies for LLM agents.