Overview

The InfluxDB OSS API v2 Query MCP Server provides a lightweight Model Context Protocol (MCP) adaptor that lets model agents and developer tools run Flux queries against an InfluxDB OSS v2 instance. It wraps the InfluxDB API behind a simple MCP-compatible HTTP interface so you can analyze, filter, and extract time‑series results programmatically from LLMs, automation agents, and other orchestration layers without embedding database credentials in each client.

This server is useful when you need a single, audited gateway between automated models and your InfluxDB deployment. It centralizes connection configuration, enforces access patterns, and returns query results in JSON that are easy to consume for downstream processing or visualization.

Features

Exposes InfluxDB query capability as an MCP tool, suitable for LLM-based agents and orchestration systems
Uses InfluxDB OSS v2 HTTP API (Flux) for full-featured time-series querying
Centralized configuration via environment variables (Influx URL, token, org, bucket)
Docker-friendly: run locally or in containers
Returns JSON query results that are easy to parse and post-process
Logging and basic request traceability for debugging and auditability

Installation / Configuration

Clone the repository, set configuration, then run locally or with Docker.

Clone the project

git clone https://github.com/idoru/influxdb-mcp-server.git
cd influxdb-mcp-server

Configure environment variables (example .env)

# .env
INFLUX_URL=https://influx.example.com      # InfluxDB OSS v2 base URL
INFLUX_TOKEN=your-influxdb-token          # Token with read/query permissions
INFLUX_ORG=your-org                       # default organization (optional)
INFLUX_BUCKET=your-bucket                 # default bucket (optional)
PORT=8080                                 # server port
LOG_LEVEL=info

Install and run (Node.js based example)

# install (if the repo uses npm)
npm install

# start
npm start
# or
node ./src/server.js

Run with Docker

docker build -t influxdb-mcp-server:latest .
docker run -e INFLUX_URL=https://influx.example.com \
           -e INFLUX_TOKEN=your-influxdb-token \
           -e INFLUX_ORG=your-org \
           -e INFLUX_BUCKET=your-bucket \
           -p 8080:8080 \
           influxdb-mcp-server:latest

Optional docker-compose snippet:

version: "3.8"
services:
  mcp:
    image: influxdb-mcp-server:latest
    environment:
      - INFLUX_URL=https://influx.example.com
      - INFLUX_TOKEN=${INFLUX_TOKEN}
      - INFLUX_ORG=your-org
      - INFLUX_BUCKET=your-bucket
    ports:
      - "8080:8080"

Environment variables reference

Variable	Required?	Purpose
INFLUX_URL	yes	Base URL for InfluxDB OSS v2 (e.g. https://influx.local:8086)
INFLUX_TOKEN	yes	Token with permissions to execute Flux queries
INFLUX_ORG	optional	Default organization to use for queries
INFLUX_BUCKET	optional	Default bucket to use for queries
PORT	optional	Port to bind the MCP server (default: 8080)
LOG_LEVEL	optional	Logging verbosity (info, debug, warn, error)

Available Resources

GitHub repository: https://github.com/idoru/influxdb-mcp-server
InfluxDB OSS v2 API docs (Flux queries): https://docs.influxdata.com/influxdb/v2.0/api-guide/
Flux language reference: https://docs.influxdata.com/flux/v0.x/

The server exposes an MCP-style tool (commonly named something like “influxdb” or “influx-query”) that accepts JSON requests containing a Flux query and optional overrides (org, bucket, time range). Results are returned as structured JSON rows with timestamps, fields and tags.

Example request pattern (MCP tool invocation)

POST /tools/influxdb
Content-Type: application/json