mcp · matih-labs — Claude Code MCP server: what it does, how to install
Links your AI assistant to the Matih data platform over MCP. Enables governed SQL execution, table profiling, and semantic metrics inside development tools.
claude mcp add --transport http mcp https://mcp.matih.ai/api/v1/mcpWhat mcp does
The server acts as a bridge between AI clients and the Matih data workspace, either over HTTP or via a local stdio process running npx -y @matihlabs/mcp. It exposes 34 stable tools alongside the experimental get_hot_context tool. The model can inspect schemas, run SQL queries, and manage dashboard charts. Metrics can also be drafted through a proposal pipeline before human review.
Access remains strictly bounded by token permissions. Running the bridge locally requires Node 20 or newer, relying solely on built-in modules without external packages. Without signed tenant egress consent, API calls will fail immediately with an EGRESS_CONSENT_REQUIRED error code.
What it helps with
- execute SQL queries against connected databases
- profile database tables and inspect column types
- draft semantic metrics for human approval
- generate dashboards and visualize query output
Example prompts for the agent
- List all tables available in the analytics warehouse
- Execute a query on the orders table and profile the status column
- Draft a new revenue metric definition based on billing data
How to install mcp
1Platform CLI: remote server
claude mcp add --transport http mcp https://mcp.matih.ai/api/v1/mcp2Platform CLI
claude mcp add --env MATIH_MCP_TOKEN=<MATIH_MCP_TOKEN> --transport stdio matih-labs-mcp -- npx -y @matihlabs/mcpReplace <…> with your own values
1Platform CLI: remote server
codex mcp add matih-labs-mcp --url https://mcp.matih.ai/api/v1/mcp2Platform CLI
codex mcp add matih-labs-mcp --env MATIH_MCP_TOKEN=<MATIH_MCP_TOKEN> -- npx -y @matihlabs/mcpReplace <…> with your own values
Config file
{
"mcpServers": {
"matih-labs-mcp": {
"url": "https://mcp.matih.ai/api/v1/mcp"
}
}
}{
"mcpServers": {
"matih-labs-mcp": {
"command": "npx",
"args": [
"-y",
"@matihlabs/mcp"
],
"env": {
"MATIH_MCP_TOKEN": "<MATIH_MCP_TOKEN>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"matih-labs-mcp": {
"type": "http",
"url": "https://mcp.matih.ai/api/v1/mcp"
}
}
}{
"servers": {
"matih-labs-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@matihlabs/mcp"
],
"env": {
"MATIH_MCP_TOKEN": "<MATIH_MCP_TOKEN>"
}
}
}
}Replace <…> with your own values
1Platform CLI: remote server
gemini mcp add -t http matih-labs-mcp https://mcp.matih.ai/api/v1/mcp2Platform CLI
gemini mcp add -e MATIH_MCP_TOKEN=<MATIH_MCP_TOKEN> matih-labs-mcp npx -y @matihlabs/mcpReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about mcp
How do I install mcp in Claude Code?
Run claude mcp add --env MATIH_MCP_TOKEN=<MATIH_MCP_TOKEN> --transport stdio matih-labs-mcp -- npx -y @matihlabs/mcp in your terminal. Replace <MATIH_MCP_TOKEN> with a developer token generated from Matih Settings.
What does the mcp server do?
It connects AI tools to Matih data services, offering governed SQL execution, table catalog discovery, and semantic model exploration.
What are the system requirements?
You need Node 20 or later to run the CLI bridge, along with a workspace token prefixed with mat_agt_.
Similar MCP servers
All MCP servers for Data and databases →mcp by other authors comes first.
