InfluxDB 3 MCP Server is a Model Context Protocol (MCP) server for InfluxDB 3 integration, officially maintained by InfluxDB. It provides tools, resources, and prompts for MCP clients to interact with InfluxDB 3 Core, Enterprise, Cloud Dedicated, Clustered, and Cloud Serverless.
Purpose The main purpose of this server is to enable AI agents or MCP clients to safely explore and query InfluxDB 3 databases. It can be used in read-only workflows, such as analysts exploring unfamiliar databases, or operators troubleshooting InfluxQL dashboard queries. By configuring read-only mode (INFLUX_MCP_TOOL_PROFILE=readonly), clients can execute only queries and data exploration without exposing write, administrative, token management, or host-level tools.
Main Capabilities The server provides three types of capabilities: tools, resources, and prompts. Available tools include list_databases, list_tables, describe_table, query_sql, query_influxql, get_help, etc., supporting structured JSON output, request and query correlation metadata (such as request_id, query_id, query_id_source), row counts, truncation status, and warning information. In read-only mode, agents can list accessible databases, discover measurements and columns, construct bounded SQL queries and execute them, while marking uncertain tag and field categories as unknown. Operators can use InfluxQL queries, SHOW queries, and schema discovery to verify measurements and columns, sample recent rows to distinguish missing data from query errors, and reject unsafe follow-up operations (such as SELECT INTO or destructive statements).
Usage Before use, environment variables must be provided, including INFLUX_DB_INSTANCE_URL, INFLUX_DB_TOKEN, and INFLUX_DB_PRODUCT_TYPE. For Cloud Dedicated, INFLUX_DB_PRODUCT_TYPE=cloud-dedicated and INFLUX_DB_CLUSTER_ID are required, along with a database token or an administrative token or both. For Clustered, INFLUX_DB_PRODUCT_TYPE=clustered and INFLUX_DB_INSTANCE_URL are required, along with the corresponding token combination. The server supports local build, npx, and Docker deployment methods. When integrating, first install dependencies (npm install), then build (npm run build), and then configure the MCP client to use the built server. Example configurations can be found in files such as example-local.mcp.json, example-npx.mcp.json, and example-docker.mcp.json. The repository also provides environment variable templates such as env.example, env.cloud-dedicated.example, env.clustered.example, and env.cloud-serverless.example.
Applicable Scenarios It is suitable for scenarios that require enabling AI agents to interact with InfluxDB 3, such as data analysis, database exploration, query debugging, dashboard troubleshooting, etc. Read-only mode is particularly suitable for exploratory analysis with high security requirements, avoiding accidental data modification.
Limitations Usage requires Node.js v20.11 or later, npm v9 or later (for the npm/npx method), and Docker (for the Docker method). Read-only mode requires setting INFLUX_MCP_TOOL_PROFILE=readonly. Tool call telemetry is enabled by default, outputting structured JSON to stderr, and does not log API tokens, request headers, tool parameters, or query text. For Core connections, resource token tools are not supported; resource tokens are supported only in Enterprise.
Other Information Information regarding pricing, charges, specific platform limitations, performance commitments, etc., is not currently provided in the available materials.
