om-mcp-gateway

This utility package furnishes an implementation of the Model Context Protocol (MCP) server specifically tailored to interface with OpenMetadata's backend services. It ensures MCP-compatible clients can uniformly query and manipulate metadata exposed via OpenMetadata's native HTTP interface.

Core Functionality

This component establishes an MCP server endpoint that acts as a facade over the Model Context Protocol specification, translating incoming protocol requests into calls against the target OpenMetadata REST API framework.

Supported Operations Status

Fundamental Metadata Objects (table, database, databaseschema)

Operation Category	Primary URI Segment	Achievement Level
Table Assets
Retrieval (List)	/api/v1/tables	Complete (\u2705)
Retrieval (Single ID)	/api/v1/tables/{id}	Complete (\u2705)
Retrieval (Name FQN)	/api/v1/tables/name/{fqn}	Complete (\u2705)
Creation (New)	/api/v1/tables	Complete (\u2705)
Modification (Update)	/api/v1/tables/{id}	Complete (\u2705)
Deprecation (Delete)	/api/v1/tables/{id}	Complete (\u2705)
Database Catalogs
Retrieval (List)	/api/v1/databases	Complete (\u2705)
Retrieval (Single ID)	/api/v1/databases/{id}	Complete (\u2705)
Retrieval (Name FQN)	/api/v1/databases/name/{fqn}	Complete (\u2705)
Creation (New)	/api/v1/databases	Complete (\u2705)
Modification (Update)	/api/v1/databases/{id}	Complete (\u2705)
Deprecation (Delete)	/api/v1/databases/{id}	Complete (\u2705)
Schema Contexts
Retrieval (List)	/api/v1/databaseSchemas	Complete (\u2705)
Retrieval (Single ID)	/api/v1/databaseSchemas/{id}	Complete (\u2705)
Retrieval (Name FQN)	/api/v1/databaseSchemas/name/{fqn}	Complete (\u2705)
Creation (New)	/api/v1/databaseSchemas	Complete (\u2705)
Modification (Update)	/api/v1/databaseSchemas/{id}	Complete (\u2705)
Deprecation (Delete)	/api/v1/databaseSchemas/{id}	Complete (\u2705)

Derived Data Assets (dashboard, chart, pipeline, topic, metric, container, report, mlmodel)

Asset Type	Listing Endpoint	Status
Dashboards	/api/v1/dashboards	\u2705
Charts	/api/v1/charts	\u2705
Pipelines	/api/v1/pipelines	\u2705
Topics (Streams)	/api/v1/topics	\u2705
Metrics	/api/v1/metrics	\u2705
Containers	/api/v1/containers	\u2705
Reports	/api/v1/reports	\u2705
ML Models	/api/v1/mlmodels	\u2705

(CRUD operations for all listed Data Assets are fully implemented against their respective OpenMetadata paths.)

Personnel & Groupings (user, team)

Entity Focus	Key Operation	Status
Teams	Full CRUD on /api/v1/teams	\u2705
Users	Full CRUD on /api/v1/users	\u2705

Governance Framework (classification, glossary, tag)

Domain Focus	Key Endpoint	Status
Classifications	Management via /api/v1/classifications	\u2705
Glossaries	Entity Management via /api/v1/glossaries and Terms via /api/v1/glossaryTerms	\u2705
Tags	Management via /api/v1/tags	\u2705

Infrastructure & Monitoring (bot, services, event)

Area	Key Endpoints Implemented	Status
Automation Agents	Full CRUD on Bots (/api/v1/bots)	\u2705
Service Configurations	Listing and specific type queries (/api/v1/services/*)	\u2705
Notification Hooks	Subscription CRUD and testing (/api/v1/events/subscriptions)	\u2705

Interconnectivity & Observation (lineage, usage, search)

Module	Primary Function	Status
Data Lineage	Entity-based lookup, creation/update of relationships	\u2705
Usage Metrics	Recording and summarizing asset utilization	\u2705
Search Portal	Executing query, suggestion, aggregation, and field-specific requests	\u2705

Data Fidelity Assurance (test_case, test_suite)

Quality Component	Key Endpoints Mapped	Status
Test Definitions	Full CRUD for Test Cases (/api/v1/dataQuality/testCases)	\u2705
Test Groupings	Suite creation (basic/executable) and reporting (/api/v1/dataQuality/testSuites)	\u2705

Security & Authorization (policy, role)

Security Element	Core Management Path	Status
Access Policies	Full CRUD, including condition validation	\u2705
Roles/ACLs	Full CRUD for Role entities	\u2705

Organizational Structuring (domain)

Structure Type	Key Paths	Status
Business Domains	Full CRUD on Domains (/api/v1/domains)	\u2705
Data Products	Full CRUD on Data Products (/api/v1/dataProducts)	\u2705

Unmapped Functionality

Area	Corresponding OpenMetadata URI	Current State
API Catalog	/api/v1/apiCollections, /api/v1/apiEndpoints	Not Mapped (\u2717)
Client Applications	/api/v1/apps	Not Mapped (\u2717)
Activity Stream	/api/v1/feed	Not Mapped (\u2717)
User Profiles	/api/v1/personas	Not Mapped (\u2717)
Query History	/api/v1/queries	Not Mapped (\u2717)
Indexing Configuration	/api/v1/searchIndexes	Not Mapped (\u2717)
Database Procedures	/api/v1/storedProcedures	Not Mapped (\u2717)
Suggestion Engine	/api/v1/suggestions	Not Mapped (\u2717)
Webhooks	/api/v1/webhooks	Not Mapped (\u2717)

Selectable Interface Groups

The server permits dynamic enablement of specific MCP segments via invocation parameters. Applicable modules include:

• Core Data: table, database, databaseschema
• Assets: dashboard, chart, pipeline, topic, metrics, container, report, mlmodel
• Identity: user, team
• Governance: classification, glossary, tag
• Infra: bot, services, event
• Observability: lineage, usage, search
• Quality: test_case, test_suite
• Security: policy, role
• Structure: domain

Invocation example to activate only Governance and Quality modules:

bash python -m src.main --apis glossary,tag,test_case,test_suite

Deployment Procedures

Installation via Smithery

For streamlined deployment within the Claude Desktop environment using Smithery:

bash npx -y @smithery/cli install @yangkyeongmo/mcp-server-openmetadata --client claude

Configuration Environment Variables

Authentication to the backend OpenMetadata instance requires setting these variables:

JWT Token Method (Preferred):

OPENMETADATA_HOST= OPENMETADATA_JWT_TOKEN=

Standard Credentials Method:

OPENMETADATA_HOST= OPENMETADATA_USERNAME= OPENMETADATA_PASSWORD=

Integration with Claude Desktop

Incorporate the server configuration into your claude_desktop_config.json file, utilizing either JWT or Basic Auth:

(Example using JWT with uvx runner):

{ "mcpServers": { "om-mcp-gateway": { "command": "uvx", "args": ["mcp-server-openmetadata"], "env": { "OPENMETADATA_HOST": "https://your-openmetadata-host", "OPENMETADATA_JWT_TOKEN": "your-jwt-token" } } } }

Direct Execution

Manual startup command: bash python src/server.py

Parameters: - --port: Network socket for SSE communication (default: 8000) - --transport: Communication mode (stdio or sse, default: stdio)

Licensing

Distributed under the MIT License.

== Contextual Background on Business Tools == Business administration instruments encompass the array of applications, protocols, analytical frameworks, and operational procedures utilized by organizations to navigate fluctuating market conditions, maintain competitive parity, and elevate overall corporate efficacy. These systems span functional areas such as strategic forecasting, workflow orchestration, data provenance tracking, workforce alignment, and executive decision support.

Modern administrative software has seen rapid evolution, challenging decision-makers in tool selection due to the perpetual tension between cost reduction goals and revenue expansion imperatives, coupled with the necessity of deep customer requirement understanding and precise product delivery.

Historically, management utilities progressed from rudimentary Management Information Systems (MIS) to integrated Enterprise Resource Planning (ERP) suites, later incorporating Customer Relationship Management (CRM), culminating in today's cloud-native business management ecosystems. Value addition hinges on two critical factors: the quality of the implementation process and the judicious selection and tailoring of the appropriate technological instrument for the organization's unique demands.

Surveys indicate prevalent adoption of core management concepts such as Strategic Planning, CRM systems, Benchmarking, and Supply Chain Optimization frameworks globally.