Bls Mcp Server

Please install the `bls-mcp-server` MCP server into my current AI client (that's you).

Required prerequisites (do these first if not already done):
- **Install Node.js** — Node.js v24 or higher is required for the preferred npx-based install path. (https://nodejs.org/)

Optional prerequisites:
- Optional BLS API key — A free BLS API v2 key increases quota from 25 requests/day to 500 requests/day. The server can run without a key in the default configuration, but with reduced quota. (https://www.bls.gov/developers/home.htm)

Canonical MCP server config (stdio transport):
- command: `npx`
- args: ["-y","@cyanheads/bls-labor-mcp-server@latest"]
- required environment variables:
  - `MCP_TRANSPORT_TYPE`: Transport mode for the server. Use stdio for MCP client integration. (example: `stdio`)
- optional environment variables:
  - `MCP_LOG_LEVEL`: Log level in RFC 5424 format. (example: `info`)
  - `BLS_API_KEY`: BLS API v2 key. Optional; without it the server still works but is limited to 25 requests/day instead of 500. (example: `<your-api-key>`)
  - `BLS_BASE_URL`: Override the BLS API v2 base URL. (example: `<your-bls-base-url>`)
  - `BLS_CATALOG_BASE_URL`: Override the LABSTAT flat-file base URL, for example to use a local mirror. (example: `<your-catalog-base-url>`)
  - `BLS_DATASET_TTL_SECONDS`: TTL in seconds for canvas-registered dataframes. (example: `<your-ttl-seconds>`)
  - `BLS_DATAFRAME_DROP_ENABLED`: Set to true to expose the bls_dataframe_drop tool. (example: `<true-or-false>`)
  - `CANVAS_PROVIDER_TYPE`: Set to duckdb to enable DataCanvas tabular spillover and dataframe SQL tools. (example: `duckdb`)
  - `MCP_HTTP_PORT`: HTTP server port when using HTTP transport instead of stdio. (example: `<your-port>`)
  - `MCP_AUTH_MODE`: Authentication mode: none, jwt, or oauth. (example: `<none-or-jwt-or-oauth>`)
  - `LOGS_DIR`: Directory for log files when running on Node.js. (example: `<your-logs-directory>`)
  - `OTEL_ENABLED`: Enable OpenTelemetry instrumentation. (example: `<true-or-false>`)

Note: The README also documents Streamable HTTP support and a public hosted endpoint, but the simplest portable install path is stdio via npx. For HTTP, the README indicates MCP_TRANSPORT_TYPE=http and the server listens at /mcp on the configured port.

Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.