Log inSubmit server
Industrial AIOps logo

Industrial AIOps

Read-first MCP access to industrial protocols with evidence-cited RCA, audit controls, and optional gated egress.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Industrial AIOps

Pick your client, then add it in one click or paste the config.

STDIOPython
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "industrial-aiops": {
      "command": "uvx",
      "args": [
        "iaiops"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

We couldn’t confirm it starts correctly

We ran the install command but it didn't respond within our test window — this can mean a slow first-time install rather than a real problem.

uvx iaiops

Process exited before answering initialize.

This is an experimental automated check and can have false negatives — missing environment variables, a slow cold install, etc. It doesn’t necessarily mean something’s wrong. Last checked 7d ago.

Overview

Industrial-AIOps MCP server reads factory-floor data across 14 field protocols and correlates it into evidence-cited operational findings. It communicates over MCP stdio, audits every call, and uses a deterministic analysis engine that does not require a model or GPU. Use it for read-first OT investigation, OEE, alarm analysis, dataflow checks, asset inventory, and governed deployments where egress must be controlled.

Use cases

  • Investigate production-line downtime across multiple OT protocols
  • Measure OEE from collected industrial history
  • Diagnose alarm floods and broken dataflows
  • Audit governed OT data access and egress
  • Check deterministic RCA results in air-gapped deployments

Key features

  • Reads 14 industrial field protocols
  • Cross-protocol root-cause analysis
  • Evidence-cited operational findings
  • Audited MCP and CLI calls
  • Deterministic, model-optional analysis
  • Air-gapped egress control

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked Today

Recent health check succeeded.

9/9 checks healthy over the last 46d

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 2d ago via OSV.dev · iaiops (PyPI)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars8Stargazers on the source repository.
Last commit2d agoMost recent push to the default branch.
Install checkInconclusiveDidn't respond in our test window — often a slow first install.
Directory activity3 viewsConfig copies, upvotes, and views on AllMCPs.

How Industrial AIOps works

What Industrial-AIOps does

Industrial-AIOps is a read-first data tap for operational technology environments. It connects to protocols including OPC UA, Modbus TCP/RTU, S7comm, Mitsubishi MC, Omron FINS, MTConnect, MQTT/Sparkplug B, EtherNet/IP, EtherCAT, PROFINET, SECS/GEM, HART-IP, BACnet/IP, and IO-Link. It also supports read-only REST access to Metasys, Niagara, and Ignition Gateway layers.

The server correlates observations across protocols instead of treating each source independently. Its analysis covers downtime root cause, ISA-18.2 alarm load, broken dataflows, data trustworthiness, OEE, asset inventory, and explanations of legacy PLC programs in ST, AWL, and L5X formats. Ten industry editions add read-only advisory checks for sectors such as factories, buildings, water, pharma, and renewables.

How it works

MCP communication uses stdio. The same governance engine is used by the MCP interface and the CLI, applying audit, budget, risk-tier, and undo controls to calls. Read paths are the default operating model; write-capable paths are limited, disabled by default, and subject to dry-run, one-shot approval, undo capture, and hash-chained audit requirements.

The analysis engine is deterministic and does not need a language model API or GPU. An optional model can rephrase an already computed and cited RCA verdict through rca_narrate; it is not responsible for the underlying calculation. Every call is audited, and the project states that it does not send telemetry or phone home.

An Industrial-AIOps MCP server can also enforce an air-gapped posture with IAIOPS_NO_EGRESS=1. That setting withholds six deliberate egress tools: stream_publish, stream_publish_event, uns_publish, historian_push, mqtt_publish, and rca_narrate.

Setup and configuration

Install the package with a protocol extra, such as pip install "iaiops[modbus]", or select [all]. Run iaiops init to create ~/.iaiops/config.yaml. Before connecting to equipment, iaiops onboard status checks local onboarding state without contacting devices, while iaiops doctor checks configuration, secrets, reachability, and version information.

The documented workflow is to create a scan plan, run the scan, draft configuration from stored scan results, collect a bounded sample, export and apply tag metadata, and then run analysis. iaiops readiness reports which scenarios are currently available and what each missing prerequisite requires. Container deployments use the published image over stdio and should mount a volume for the audit store. The repository also provides hardened, air-gapped deployment material.

Tools and capabilities

  • Survey devices with iaiops scan plan and iaiops scan run.
  • Collect time-bounded data with iaiops collect run.
  • Measure OEE with iaiops oee measure.
  • Open investigations with iaiops investigate open and run RCA with iaiops diag rca.
  • Export and apply tag metadata with iaiops tags export and iaiops tags apply.
  • Validate deterministic analysis with iaiops verify determinism; the MCP-side equivalent is verify_determinism.
  • Use protocol-specific reads, cross-protocol correlation, audit controls, and selected governed egress operations.

Limitations and notes

The documented verification covers protocol libraries, containers, and in-process servers, but not real plant equipment. Evidence levels differ by protocol: some use third-party implementations, some use project-written servers, and some are mock-only. The separate iaiops-energy project contains substation and utility telecontrol support for IEC-104, DNP3, and IEC-61850; those protocols are not part of this package.

Although the project describes a read-first design, several tools can send data or perform other externally visible actions. Review the audit, approval, risk, and egress controls before enabling them in an OT environment. The optional narration model does not replace the deterministic, cited result.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Industrial AIOps

Industrial-AIOps is an MCP server that connects AI agents to industrial field protocols and read-only factory data layers. Its main tools collect and inspect OT data, measure OEE, investigate downtime, analyze alarms and dataflows, run cross-protocol RCA, and verify deterministic results with audited controls.

Install it with a protocol-specific extra such as `pip install "iaiops[modbus]"`, or use `[all]`. Then run `iaiops init` to create the local configuration, followed by `iaiops onboard status` and `iaiops doctor` to inspect onboarding and setup before contacting equipment.

Yes, Industrial-AIOps supports an air-gapped posture. Set `IAIOPS_NO_EGRESS=1` to withhold its six deliberate egress tools, and use the repository's air-gap deployment material. The core analysis engine requires neither a model API nor a GPU.

Industrial-AIOps supports OPC UA, Modbus TCP/RTU, S7comm, Mitsubishi MC, Omron FINS, MTConnect, MQTT/Sparkplug B, EtherNet/IP, EtherCAT, PROFINET, SECS/GEM, HART-IP, BACnet/IP, and IO-Link. It also reads selected Metasys, Niagara, and Ignition Gateway REST layers.

The project reports verification against real protocol libraries, containers, and in-process servers, but not real plant gear. The evidence level varies by protocol, including third-party servers, project-written servers, and mock-only coverage.

Related MCP servers

All in Industrial & IoT All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewIndustrial AIOps AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/industrial-aiops?style=directory)](https://allmcps.com/mcp/industrial-aiops)
HTML Embed
<a href="https://allmcps.com/mcp/industrial-aiops"><img src="https://allmcps.com/api/badge/industrial-aiops?style=directory" alt="Industrial AIOps on AllMCPs" /></a>