MyScale (SQL Vector Database API)

MyScale (SQL Vector Database API) MCP Connector for Claude

F

Manage MyScale vector databases via SQL — execute queries, perform vector searches, and manage indices directly from any AI agent.

6 tools Official Updated Jun 28, 2026 Official Vinkius Partner

Connect your MyScale cluster to any AI agent to manage high-performance vector data using familiar SQL syntax. MyScale combines the power of ClickHouse with advanced vector search capabilities.

What you can do

  • SQL Execution — Run arbitrary SQL queries including SELECT, INSERT, and ALTER directly on your cluster via execute_sql_query.
  • Vector Search — Perform similarity searches using the vector_search tool which automatically handles distance functions and SQL construction.
  • Index Management — Create and monitor specialized vector indices (MSTG, ScaNN, HNSW) using create_vector_index and check_index_status.
  • Schema Control — Quickly set up new environments with create_vector_table, defining dimensions and constraints through natural language.
  • Connectivity Checks — Instantly verify your database status and credentials with ping_cluster.

How it works

  1. Subscribe to this server
  2. Enter your MyScale Cluster URL, Username, and Password
  3. Start managing your vector data from Claude, Cursor, or any MCP-compatible client

Who is this for?

  • AI Engineers — test vector search parameters and inspect data distributions without leaving the chat
  • Data Scientists — manage RAG (Retrieval-Augmented Generation) backends and monitor index health in real-time
  • Backend Developers — perform database maintenance and schema updates using natural language commands
sqlvector-searchragclickhousedatabase-management

6 tools expose this connector's capabilities to your AI agent.

check_index_status

vector_indices table to monitor the build status of vector indices (Built, InProgress, Error). Check the status of vector indices

create_vector_index

ADD VECTOR INDEX statement. Create a vector index on a table

create_vector_table

Create a new table with a vector column

execute_sql_query

For SELECT queries, FORMAT JSON is automatically appended. Execute an arbitrary SQL query on MyScale

ping_cluster

Check MyScale cluster connectivity

vector_search

Automatically constructs the SQL query. Perform a vector similarity search

See how to talk to your AI agent using MyScale (SQL Vector Database API).

Check if my MyScale cluster is reachable and credentials are correct.

I've pinged the cluster using `ping_cluster`. The connection is successful and the database is responding normally.

Search the 'articles' table for the top 5 matches for vector [0.12, -0.05, 0.33] where author is 'John'.

I'm executing a `vector_search` on the 'articles' table. I've applied the filter for author 'John' and limited the results to 5. Here are the most relevant documents found...

Create a new vector table called 'product_embeddings' with 1536 dimensions and an 'id' column.

I've initiated the `create_vector_table` tool for 'product_embeddings'. The table has been created with a 1536-dimension float array constraint and an additional 'id' column using the MergeTree engine.

Use the `check_index_status` tool. It queries the system tables to show you the current status (Built, InProgress, or Error) for all vector indices in your cluster.

Related Connectors