# jovezhong/mcp-timeplus [Health: Active]

**Category:** 📊 Data Platforms  
**Repository:** https://github.com/jovezhong/mcp-timeplus  
**GitHub Stars:** 12  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/jovezhong-mcp-timeplus

## Description
MCP server for Apache Kafka and Timeplus. Able to list Kafka topics, poll Kafka messages, save Kafka data locally and query streaming data with SQL via Timeplus

## Tools
Capabilities this server exposes over MCP:

- **run_sql**
- **sql**
- **TIMEPLUS_READ_ONLY**
- **false**
- **list_databases**
- **list_tables**
- **database**
- **list_kafka_topics**
- **explore_kafka_topic**
- **topic**
- **message_count**
- **create_kafka_stream**
- **connect_to_apache_iceberg**
- **iceberg_db**
- **aws_account_id**
- **s3_bucket**
- **aws_region**
- **is_s3_table_bucket**

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `uvx` (confidence: high):

```json
"mcpServers": {
  "mcp-timeplus": {
    "command": "uvx",
    "args": ["mcp-timeplus"],
    "env": {
      "CLICKHOUSE_HOST": "",
      "CLICKHOUSE_USER": "",
      "CLICKHOUSE_PASSWORD": ""
    }
  }
}
```

**Requires environment variables:** `CLICKHOUSE_HOST`, `CLICKHOUSE_USER`, `CLICKHOUSE_PASSWORD` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What jovezhong/mcp-timeplus MCP server does

jovezhong/mcp-timeplus MCP server exposes database and SQL operations to an MCP client. The README identifies the implementation as a ClickHouse MCP server and documents connections to a ClickHouse cluster.

The documented workflow is oriented toward inspection and querying rather than data mutation. SQL sent through `run_select_query` is executed with ClickHouse `readonly = 1`, which limits the documented query path to read-only operations.

## How it works

An MCP client starts the server with `uv run`, supplies connection settings through environment variables, and invokes the exposed tools. The documented tools are:

- `run_select_query`: executes a SQL statement against the configured ClickHouse cluster.
- `list_databases`: returns the databases available on that cluster.
- `list_tables`: returns tables for a database supplied as an input.

The supplied tool inventory also lists `sql` and `database`.

## Setup and configuration

The command uses `uv` with the `mcp-clickhouse` package and Python 3.13. Replace the placeholder connection values with those for the target ClickHouse service, then restart Claude Desktop.

`CLICKHOUSE_HOST`, `CLICKHOUSE_USER`, and `CLICKHOUSE_PASSWORD` are required by the documented configuration. `CLICKHOUSE_PORT`, `CLICKHOUSE_SECURE`, and `CLICKHOUSE_VERIFY` control the endpoint and TLS behavior. `CLICKHOUSE_CONNECT_TIMEOUT` and `CLICKHOUSE_SEND_RECEIVE_TIMEOUT` set connection and query communication limits, while `CLICKHOUSE_DATABASE` selects a default database. The documented defaults use port 8443 for secure connections and 8123 for non-secure connections.

## Limitations and notes

The repository name and listing refer to jovezhong/mcp-timeplus MCP server, Apache Kafka, and Timeplus, while the README title, package name, configuration variables, and documented tools refer to ClickHouse. Confirm the repository’s current implementation before depending on Kafka or Timeplus-specific behavior.

_Full upstream README: https://allmcps.com/mcp/jovezhong-mcp-timeplus/readme_