Provides a comprehensive set of tools for interacting with MongoDB databases, including full CRUD operations on documents, database and collection management, index creation and maintenance, and complex aggregation pipeline execution.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Mongo-MCPfind documents in the 'users' collection where status is 'active'"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Mongo-MCP
English | 简体中文
A Machine Chat Protocol (MCP) service for MongoDB operations. This service provides a comprehensive set of tools that allow Large Language Models (LLMs) to interact with MongoDB databases through complete CRUD operations, administrative tasks, and advanced features.
Requirements
Python 3.10 or above
A running MongoDB database service
It is recommended to use uv to run the program
🚀 Features
📊 Database Management Tools
list_databases- List all databasescreate_database- Create new databasedrop_database- Delete databaseget_database_stats- Get database statistics
📦 Collection Management Tools
list_collections- List all collections in a databasecreate_collection- Create new collection (with optional settings)drop_collection- Delete collectionrename_collection- Rename collectionget_collection_stats- Get collection statistics
📄 Document CRUD Operations
insert_document- Insert single documentinsert_many_documents- Batch insert multiple documentsfind_documents- Query documents (supports sorting, projection, limit)find_one_document- Query single documentcount_documents- Count documents matching queryupdate_document- Update documents (single or batch)replace_document- Replace documentdelete_document- Delete documents (single or batch)
🔍 Index Management Tools
list_indexes- List all indexes for a collectioncreate_index- Create regular indexcreate_text_index- Create text search indexcreate_compound_index- Create compound indexdrop_index- Delete indexreindex_collection- Rebuild all indexes for a collection
📈 Aggregation Operations
aggregate_documents- Execute aggregation pipeline operationsdistinct_values- Get distinct values for a field
🔧 Monitoring and Administrative Tools
get_server_status- Get MongoDB server statusget_replica_set_status- Get replica set statusping_database- Test database connectiontest_mongodb_connection- Comprehensive connection testget_connection_details- Get detailed connection information
🛠️ Technology Stack
Python: Primary programming language
FastMCP: MCP Python SDK for automatic tool definition generation
PyMongo: Official MongoDB Python driver
uv: Modern Python package management tool
Usage
Run directly with uvx
The server uses the stdio transport method, making it suitable for integration with MCP clients that support this transport method.
Cursor Example Configuration
If you use Cursor as your development environment, you can add the following configuration to your .cursor/mcp.json file for local debugging:
Environment Variables
Basic Configuration
MONGODB_URI: MongoDB connection string (default: "mongodb://localhost:27017")MONGODB_DEFAULT_DB: Default database name (optional)
Connection Pool Configuration
MONGODB_MIN_POOL_SIZE: Minimum connection pool size (default: 0)MONGODB_MAX_POOL_SIZE: Maximum connection pool size (default: 100)MONGODB_MAX_IDLE_TIME_MS: Maximum idle time in milliseconds (default: 30000)
Timeout Configuration
MONGODB_SERVER_SELECTION_TIMEOUT_MS: Server selection timeout (default: 30000)MONGODB_SOCKET_TIMEOUT_MS: Socket timeout (default: 0 - no timeout)MONGODB_CONNECT_TIMEOUT_MS: Connection timeout (default: 20000)
Security Configuration
MONGODB_TLS_ENABLED: Enable TLS connection (default: false)MONGODB_AUTH_SOURCE: Authentication source (default: admin)MONGODB_AUTH_MECHANISM: Authentication mechanism (SCRAM-SHA-1, SCRAM-SHA-256, etc.)
Performance Settings
MONGODB_READ_PREFERENCE: Read preference (default: primary)MONGODB_WRITE_CONCERN_W: Write concern (default: 1)MONGODB_READ_CONCERN_LEVEL: Read concern level (default: local)
Logging Configuration
LOG_LEVEL: Logging level (default: "INFO")Available values: DEBUG, INFO, WARNING, ERROR, CRITICAL
LOG_MAX_FILE_SIZE: Maximum log file size in bytes (default: 10MB)LOG_BACKUP_COUNT: Number of backup log files (default: 5)
Feature Flags
ENABLE_DANGEROUS_OPERATIONS: Enable potentially dangerous operations (default: false)ENABLE_ADMIN_OPERATIONS: Enable administrative operations (default: true)ENABLE_INDEX_OPERATIONS: Enable index operations (default: true)
Development Guide
Clone the repository
Install development dependencies
Run tests
Code Structure
src/mongo_mcp/server.py: MCP server implementationsrc/mongo_mcp/db.py: Core MongoDB operations implementationsrc/mongo_mcp/config.py: Configuration managementsrc/mongo_mcp/tools/: MCP tools implementationdatabase_tools.py: Database and collection managementdocument_tools.py: Document CRUD operationsindex_tools.py: Index managementaggregation_tools.py: Aggregation operationsadmin_tools.py: Administrative and monitoring tools
src/mongo_mcp/utils/: Utility modulestests/: Test cases
Testing
The project includes comprehensive test coverage:
Unit tests for all tool modules
Integration tests with MongoDB
Mock tests for isolated component testing
Run the test suite:
Logging
Log files are stored in the logs directory by default. The logging system supports:
Configurable log levels
File rotation based on size
UTF-8 encoding support
Structured logging with function names and line numbers
License
MIT
Contributing
Contributions via Issues and Pull Requests are welcome. Before submitting a PR, please ensure:
All tests pass (
uv run pytest)Appropriate test cases are added
Documentation is updated
Code follows the existing style patterns