# ober37/ac-infinity-mcp [Health: Active]

**Category:** 🏭 Industrial & IoT  
**Repository:** https://github.com/ober37/ac-infinity-mcp  
**GitHub Stars:** 9  
**Views:** 4  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/ober37-ac-infinity-mcp

## Description
Monitor and automate AC Infinity grow controllers through natural conversation with Claude. 25 tools covering live sensor data, multi-day history, VPD/temperature/humidity automations, port control, advance scheduling, and grow stage templates.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `uvx` (confidence: high):

```json
"mcpServers": {
  "ac-infinity-mcp": {
    "command": "uvx",
    "args": ["--from","git+https://github.com/ober37/ac-infinity-mcp.git","ac-infinity-mcp"],
    "env": {
      "AC_INFINITY_EMAIL": "",
      "AC_INFINITY_PASSWORD": ""
    }
  }
}
```

**Requires environment variables:** `AC_INFINITY_EMAIL`, `AC_INFINITY_PASSWORD` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What ober37/ac-infinity-mcp MCP server does

ober37/ac-infinity-mcp MCP server exposes AC Infinity controller data and controls through the Model Context Protocol. It connects to the AC Infinity cloud API, allowing an MCP client to discover devices, inspect sensor readings, review port state, and request controller changes without using the AC Infinity mobile application directly.

Supported hardware listed by the project includes the UIS Controller 69 Pro, UIS Controller 69 Pro+, and UIS Controller 89 AI+. The 69 Pro models support reads and writes through the legacy protocol. The 89 AI+ supports readings and per-port writes, with protocol-specific requirements described in the project documentation.

## How it works

The server runs as a Python application built with FastMCP and communicates with AC Infinity using account credentials supplied at runtime. An MCP client starts the server over stdio and presents its tools and prompts to the assistant. The assistant can then call read-only operations for current or historical data, or prepare and execute device changes.

Write tools default to `dry_run=True`. In that mode, the server returns the payload it would send instead of changing equipment. A caller must explicitly use `dry_run=False` to perform a write. This applies to port controls, automation settings, grow-stage templates, and Advance Automation operations.

Three prompts are also registered: one for VPD troubleshooting, one for setting up a new grow, and one for interpreting environment alerts and health grades.

## Setup and configuration

Install with pip from the Git repository, create an isolated virtual environment, or run the executable with `uvx`. Python 3.11 or later is required. The package provides an `ac-infinity-mcp` executable for MCP client configuration.

Set these required environment variables:

- `AC_INFINITY_EMAIL`: AC Infinity account email
- `AC_INFINITY_PASSWORD`: AC Infinity account password

Claude Desktop, Cursor, and Windsurf examples in the project use a stdio command configuration with these variables in the client environment. Docker Compose is also documented; the container runs over stdio and can be connected through a stdio bridge such as `mcp-proxy`.

## Tools and capabilities

The ober37/ac-infinity-mcp MCP server includes discovery, readings, analytics, port status, write, automation, grow-stage, and Advance Automation tools. Read operations include device discovery, current temperature, humidity, VPD, port states, historical time-series data, port settings, and port activity reports.

Analytics tools can check VPD drift for a grow stage, calculate an environmental health score from 0 to 100 with an A–F grade, and identify linear trends with a seven-day projection. Port controls can set speed from 1 to 10, turn ports on or off, switch modes, and configure temperature, humidity, or VPD automation.

Advance Automation tools can list, inspect, create, enable, disable, update, and delete named programs. They can also add or remove individual rules. Grow-stage templates apply combined VPD, temperature, and humidity automation, while the breakout operation returns a safe way to remove a port from automation control.

## Limitations and notes

AI+ controller support is not uniform across every automation feature. Reads and some in-place Advance Automation edits are documented as verified, while creation, enable/disable behavior, and other operations have limited or hardware-specific verification. `apply_grow_stage_template` and `break_out_of_automation` are held back on AI+ pending hardware verification and refuse to write there.

The project notes that per-port writes on the 89 AI+ require a `minversion` header. The upstream AC Infinity cloud API does not support HTTPS, so deployment should account for that transport limitation when handling account credentials and network traffic.

_Full upstream README: https://allmcps.com/mcp/ober37-ac-infinity-mcp/readme_