Skip to content
InfluxDB 3 MCP Server

InfluxDB 3 MCP Server

by InfluxData
Added to Onei
v1.4.0
Apache-2.0 OR MIT
Repository Docs npm
Featured Databases & Data

Summary

InfluxData's official MCP server for InfluxDB 3 — schema discovery, bounded SQL and InfluxQL queries, line-protocol writes and token administration across Core, Enterprise and Cloud.

Features

  • One server across InfluxDB 3 Core, Enterprise, Cloud Serverless, Cloud Dedicated and Clustered
  • Read-only tool profile that hides every write, admin and token tool
  • Bounded SQL and InfluxQL queries with row counts, truncation flags and correlation IDs
  • Schema discovery via list_tables, describe_table and investigate_database
  • Line-protocol writes and full database lifecycle management
  • Admin and resource token management on Core and Enterprise
  • MCP resources and prompts, plus a loadable custom database context file

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

  1. Open your agent's MCP configuration file.
  2. Merge the snippet below into it, filling in the environment variables with your own values.
  3. Restart the agent — the "InfluxDB 3 MCP Server" tools will be available.
{
  "mcpServers": {
    "influxdb3-mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@influxdata/influxdb3-mcp-server"
      ],
      "env": {
        "INFLUX_DB_INSTANCE_URL": "http://localhost:8181/",
        "INFLUX_DB_TOKEN": "<YOUR_INFLUXDB_TOKEN>",
        "INFLUX_DB_PRODUCT_TYPE": "core"
      }
    }
  }
}

Description

InfluxData's official Model Context Protocol server for InfluxDB 3. It exposes one connection surface across every InfluxDB 3 deployment — Core, Enterprise, Cloud Serverless, Cloud Dedicated and Clustered — and adapts the available tools to whichever product type you configure, so Core/Enterprise token management and Cloud Dedicated database-token tools only appear where they apply.

Read-only by default is a first-class mode. Setting INFLUX_MCP_TOOL_PROFILE=readonly removes every write, admin, token-management and host-level tool from the agent's view. What remains is a genuinely useful exploration loop: list_databases to see what is reachable, list_tables and describe_table to discover measurements and columns, then query_sql or query_influxql with bounded results and structured response metadata. That is the profile to hand an analyst agent or anything pointed at production.

Queries come back with provenance. Responses carry row count, truncation status, warnings, request_id, query_id and query_id_source, so a result can be correlated against system.queries.id when query history is available. Unsafe follow-ups such as SELECT INTO or destructive InfluxQL statements are rejected rather than executed. Structured logs go to stderr, keeping stdout clean for MCP protocol messages over stdio.

Full surface. Beyond querying: write_line_protocol for ingest, database create/update/delete with product-specific configuration such as retention and Cloud Dedicated maxTables/maxColumns, investigate_database for high-level discovery and sampling, health_check, plus admin and resource token lifecycle on Core and Enterprise. It also publishes MCP resources (influx-config, influx-status, influx-databases, context-file) and prompts, and supports loading a custom context file so the agent knows your schema conventions before it writes a query.

Runs via npx, a local Node build, or Docker. Requires Node.js 20.11+ and npm 9+. Dual-licensed Apache-2.0 OR MIT.

Related MCP Servers

MCP: PlanetScale

by PlanetScale

PlanetScale's official hosted MCP server — browse organizations, databases and branches, read schema, run guarded SQL, and pull Insights and schema recommendations.

Databases & Data
Browse all MCP servers →