Skip to main content
MCP server CX Codex

InfluxDB 3 — Claude Code MCP server: query and write databases

The official InfluxDB 3 MCP server lets an agent read and write data, inspect schemas and trace query errors. Handy when you want to dig into a database from chat.

Install into Claude Code · Platform CLI All methods and platforms ↓
claude mcp add --env INFLUX_DB_PRODUCT_TYPE=<INFLUX_DB_PRODUCT_TYPE> --env INFLUX_DB_TOKEN=<INFLUX_DB_TOKEN> --env INFLUX_DB_INSTANCE_URL=<INFLUX_DB_INSTANCE_URL> --env INFLUX_DB_MANAGEMENT_TOKEN=<INFLUX_DB_MANAGEMENT_TOKEN> --env INFLUX_DB_ACCOUNT_ID=<INFLUX_DB_ACCOUNT_ID> --transport stdio influxdb-3 -- npx -y @influxdata/influxdb3-mcp-server
Open on GitHub

What InfluxDB 3 does

The server hands your agent tools for InfluxDB 3 Core, Enterprise, Cloud Dedicated, Clustered and Cloud Serverless. It can run SQL and InfluxQL, write points, and manage databases and tokens. Which tools show up depends on the profile: with INFLUX_MCP_TOOL_PROFILE=readonly only read and discovery tools remain, and anything that mutates data or administers the instance is left out.

The token travels in a Bearer header and the server never inspects it, so the token's own permissions decide what is reachable. A scoped token simply won't see the rest. Failures come back structured, with error.code, error.recovery and error.retryable. A 403 maps to not_found_or_denied, and a query against a table that doesn't exist tells you to fix the reference via list_tables or describe_table.

You need Node.js v20.11 or newer plus the URL and token of an InfluxDB 3 instance. The Docker path needs Docker. Local development runs on pnpm, pinned in package.json and activated with corepack enable.

What it helps with

  • run SQL or InfluxQL against a database from inside the agent
  • list databases, tables and columns before writing a query
  • figure out why a dashboard panel went blank after a deploy
  • tell missing data apart from a broken query
  • give an agent read access with no write or admin tools

Example prompts for the agent

  • List the databases and tables in InfluxDB, then average the sensor temperature over the last day
  • Why does this InfluxQL query return nothing? Check the measurement and its columns
  • Build a bounded SQL query against sensors and return the rows with any warnings

How to install InfluxDB 3

Platform CLI

claude mcp add --env INFLUX_DB_PRODUCT_TYPE=<INFLUX_DB_PRODUCT_TYPE> --env INFLUX_DB_TOKEN=<INFLUX_DB_TOKEN> --env INFLUX_DB_INSTANCE_URL=<INFLUX_DB_INSTANCE_URL> --env INFLUX_DB_MANAGEMENT_TOKEN=<INFLUX_DB_MANAGEMENT_TOKEN> --env INFLUX_DB_ACCOUNT_ID=<INFLUX_DB_ACCOUNT_ID> --transport stdio influxdb-3 -- npx -y @influxdata/influxdb3-mcp-server

Replace <…> with your own values. 2 more environment variables are in the server README

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 InfluxDB 3

How do I install InfluxDB 3 in Claude Code?

Run `claude mcp add influxdb3-mcp-server -e INFLUX_DB_PRODUCT_TYPE=<INFLUX_DB_PRODUCT_TYPE> -e INFLUX_DB_TOKEN=<INFLUX_DB_TOKEN> -e INFLUX_DB_MANAGEMENT_TOKEN=<INFLUX_DB_MANAGEMENT_TOKEN> -- npx -y @influxdata/influxdb3-mcp-server`. Replace the placeholders with your own values.

What does the InfluxDB 3 MCP server do?

It gives an agent tools to read and write InfluxDB 3 data, explore schemas and diagnose query errors. In read-only mode the write and admin tools are dropped.

What do I need to run the server?

An InfluxDB 3 instance with its URL and token, plus Node.js v20.11 or newer. The Docker setup also needs Docker.

Similar MCP servers

All MCP servers for Data and databases →

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor