Loading...
This MCP server provides access to Milvus vector database functionalities, enabling LLM applications to perform vector search, data management, and collection operations.
Boost this tool
Subscribe to listing upgrades or segmented pushes.
This MCP server provides access to Milvus vector database functionalities, enabling LLM applications to perform vector search, data management, and collection operations.
This MCP server offers a range of tools, including potentially destructive operations like deleting entities. While authentication is supported, the lack of sandboxing and direct database access necessitate careful configuration and usage. It is safe for read-only operations and collection listing, but risky when performing data deletion or modification without proper controls.
Performance depends on the size of the Milvus collection, indexing strategy, and hardware resources. Consider using appropriate indexing techniques and scaling the Milvus cluster as needed.
Cost depends on the Milvus deployment (cloud or on-premise), resource consumption, and data storage requirements. Factor in the cost of API calls and data transfer.
{
"mcpServers": {
"milvus-sse": {
"url": "http://your_sse_host:port/sse",
"disabled": false,
"autoApprove": []
}
}
}milvus_text_searchSearches for documents within a specified Milvus collection using full-text search capabilities.
Read-only operation; retrieves data based on a text query.
milvus_vector_searchPerforms vector similarity search on a Milvus collection using a provided query vector.
Read-only operation; retrieves similar vectors based on a given query vector.
milvus_hybrid_searchExecutes a hybrid search on a Milvus collection, combining text and vector queries for enhanced results.
Read-only operation; combines text and vector queries to retrieve data.
milvus_text_similarity_searchPerforms text similarity search on a Milvus collection, leveraging embedding functions for semantic matching.
Read-only operation; uses embedding functions to find semantically similar text.
milvus_queryQueries a Milvus collection using a filter expression to retrieve specific data.
Read-only operation; retrieves data based on a specified filter expression.
milvus_list_collectionsLists all collections available in the Milvus database.
Read-only operation; retrieves a list of collection names.
milvus_create_collectionCreates a new collection in the Milvus database with a specified schema.
Creates a new collection, modifying the database structure.
milvus_load_collectionLoads a Milvus collection into memory for faster search and query operations.
Loads a collection into memory, potentially impacting resource usage.
milvus_release_collectionReleases a Milvus collection from memory to free up resources.
Releases a collection from memory, potentially impacting performance.
milvus_get_collection_infoRetrieves detailed information about a specific Milvus collection, including schema and metadata.
Read-only operation; retrieves collection metadata.
milvus_insert_dataInserts data into a Milvus collection.
Writes data into a collection, modifying its content.
milvus_delete_entitiesDeletes entities from a Milvus collection based on a filter expression.
Deletes data from a collection, potentially causing irreversible data loss.
Token|Environment Variable
hybrid
This MCP server offers a range of tools, including potentially destructive operations like deleting entities. While authentication is supported, the lack of sandboxing and direct database access necessitate careful configuration and usage. It is safe for read-only operations and collection listing, but risky when performing data deletion or modification without proper controls.
Autonomy level depends on the configured permissions and the specific tools used. Exercise caution when enabling destructive tools in autonomous workflows.
Production Tip
Monitor Milvus resource usage and performance to ensure optimal operation in production environments.
MCP is an open protocol that enables seamless integration between LLM applications and external data sources and tools.
Milvus is a vector database designed for storing, indexing, and managing massive embedding vectors.
You need to modify the `claude_desktop_config.json` file to specify the server command and arguments or the SSE endpoint.
The server supports stdio mode (default) and SSE (Server-Sent Events) mode.
Run the server with the `--sse` flag and specify the port using the `--port` flag.
The server supports `MILVUS_URI`, `MILVUS_TOKEN`, and `MILVUS_DB` environment variables.
Check if 'milvus' or 'milvus-sse' appear in the list in Cursor Settings > MCP and confirm that the relevant tools are listed.