Predictive Maintenance MCP plugin
A Claude Code plugin and MCP server for machine vibration diagnostics: fault frequencies, ISO 20816-3 severity and reports.
- 114 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The server reads local signal and manual files and writes reports to disk, and the README shows no network actions.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/predictive-maintenance-mcpDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The server reads local signal and manual files and writes reports to disk, and the README shows no network actions.
This is third-party code. Review the repository files before installing.
What it does
An MCP server for reliability engineers. The agent loads a vibration signal, tests the envelope spectrum for characteristic bearing fault frequencies and rates severity under ISO 20816-3. The server refuses to guess: a fault claim needs matching spectral lines, and the evaluative wording in reports is written by the server, not the model. A Claude Code plugin ships with nine skills (bearing and gear diagnosis, quick screening, reports, anomaly detection, prognostics), two agents and the `/pm-diagnose`, `/pm-screen`, `/pm-report` commands. There are 48 entry points, 44 tools and 4 prompts, with 20 sample signals in the clone.
Who it is for. Reliability and maintenance engineers who analyze bearing and gear vibration with an agent.
Good fit when
- A bearing must be checked from a vibration recording with matched fault frequencies.
- You need the ISO 20816-3 severity zone for a machine group.
- Bearing data must be extracted from a machine manual PDF to compute expected frequencies.
- Fresh pump measurements are compared with a declared baseline.
Not a fit when
- You expect a final verdict without engineer review: the final call stays with the human.
- Data is not a signal recording with a known sampling rate.
- You need real-time online monitoring wired to sensors.
Example request
Load `real_train/OuterRaceFault_1.csv`, shaft at 1500 rpm: check the bearing against the metadata fault frequencies, assess the ISO zone and generate a report.Limitations
Conclusions rest on the envelope spectrum and may fit more than one fault type, so the server lists them without picking one. For your own data you need a headerless CSV and a `<name>_metadata.json` file with `sampling_rate` and `signal_unit`. Sample signals ship with the clone, not the PyPI package. Some tools from the Unreleased section of the changelog are not yet in the published package. Check the claimed accuracy and benchmark against the methodology document.
How to disable. Remove the plugin with `/plugin uninstall` and the `predictive-maintenance` entry from the MCP client settings. The `reports` and `data/signals` folders in the clone hold your data, clean them by hand.
FAQ
How do I add the plugin to Claude Code?
Run `/plugin marketplace add LGDiMaggio/predictive-maintenance-mcp`, then `/plugin install predictive-maintenance@predictive-maintenance-marketplace`.
How do I connect the server to Claude Desktop?
Clone the repository, create an environment with `uv`, install the package and add a `predictive-maintenance` entry to `claude_desktop_config.json` with absolute paths. On Windows `setup_claude.ps1` does this.
What about my own data?
Put a single-column headerless CSV into `data/signals/` and add a `<name>_metadata.json` next to it. Then ask for the analysis with the shaft speed.
Does the server read the diagnosis from the file name?
No, file names are ignored. The conclusion rests on the spectrum alone.
Which license?
MIT.
Discussion
No questions or reviews yet.
Related
PluginsMonte Carlo plugin for Claude Code and Cursor: table checks before queries, data incident triage and monitor creation.
An MCP server with a toolbox for global briefs: markets, SEC filings, conflict, climate and news from open sources
Official Pinecone plugin: an agent creates indexes, searches them, builds a RAG assistant and cites your files.
SP-Mind is a Claude Agent SDK agent for spatial proteomics analysis: from image correction to cell type annotation.
A Claude Code agent for protein design: target research, antibody and nanobody generation, screening and candidate ranking.
MySQL MCP Server Pro
mysql_mcp_server_pro
A MySQL MCP server that runs SQL under access roles, finds tables by comments and analyzes locks, indexes and query plans
README in short
The README describes the no-guessing rule of the server, sample requests, a quick start for Claude Desktop, the Claude Code plugin, a benchmark and paths for engineers and developers. It links to `INSTALL.md` and the methodology.