Overview
Features
Atlas API access with service accounts
Authenticate against Atlas using MDB_MCP_API_CLIENT_ID and MDB_MCP_API_CLIENT_SECRET to enable Atlas tooling and API interactions.
Atlas Tools suite
Provides Atlas-related MCP tools (atlas-list-orgs, atlas-list-projects, atlas-create-project, atlas-list-clusters, etc.) for managing Atlas resources.
MongoDB Database Tools
Includes tools like connect, find, aggregate, count, insert-many, create-index, update-many, etc., for data access and manipulation.
Flexible configuration methods
Configure the server via environment variables, MCP configuration files, command-line arguments, and optional proxies.
HTTP transport support
Run the MCP server as an HTTP server using --transport http with configurable httpHost and httpPort.
Logging and telemetry
Configurable loggers (disk, mcp, stderr), logPath, and telemetry control.
Preview features: Vector Search
Preview features such as vectorSearch enabled via Voyage AI embedding models, requiring voyageApiKey.
Exports management
Data exports are managed via exportsPath with exportTimeoutMs and exportCleanupIntervalMs, exposing an exported-data resource.
Who Is This For?
- Developers:Build MCP-enabled apps to access and manage MongoDB Atlas and databases.
- DBAs/DevOps:Coordinate Atlas and MongoDB data operations across cloud and on-prem deployments.
- MCP Integrators:Develop MCP clients to query, monitor, and manage MongoDB data.




