Ramly for AI assistants

Run reliability, availability and maintainability (RAM) studies from Claude, ChatGPT, Microsoft Copilot, Cursor, VS Code and other AI tools with Ramly.

You describe a plant, process or fleet. The assistant:

builds the reliability block diagram;
runs Ramly's discrete-event Monte Carlo or exact analytical (Markov) engine on your Ramly account;
explains availability, downtime, bad actors, spares, crews and costs;
compares design options.

This repository contains:


`skills/ramly-ram-analysis/`	An Agent Skill with the RAM study workflow, modeling patterns, distribution guidance, result interpretation and worked examples
`.mcp.json` / `mcp.json`	The Ramly MCP server, `https://mcp.ramly.io/mcp` (OAuth sign-in to your Ramly account)
`.claude-plugin/`	Claude plugin manifest and marketplace
`plugin.json`	Agent Plugins 1.0 manifest (Codex, Copilot, Cursor and others)

A Ramly account is required. Studies use your plan's monthly run credits. Plans are at ramly.io.

Install

Claude Code

/plugin marketplace add arslanerdem/ramly-plugin
/plugin install ramly@ramly

The first time a Ramly tool runs, Claude opens a browser to sign in to Ramly.

Claude (web, desktop, mobile)

Go to Settings → Connectors → Add custom connector.
Enter the URL https://mcp.ramly.io/mcp.
To add the skill, upload skills/ramly-ram-analysis as a ZIP under Settings → Capabilities → Skills.

ChatGPT, Microsoft Copilot Studio, VS Code, Cursor and other MCP clients: add an MCP server or connector with the URL https://mcp.ramly.io/mcp and sign in when asked. Clients that support Agent Plugins or Agent Skills can install this repository directly.

What you can ask

"Our cooling water system has three pumps (2 needed) and a heat exchanger. Model it in Ramly and tell me the availability."
"Would a second spare pump cartridge pay for itself?"
"Which components drive our downtime? Show the importance ranking."
"What PFD and SIL does our 1oo2 high-pressure trip reach with annual proof testing?"
"Compare the duty/standby design with 2 × 100 % hot redundancy."

Data and privacy

The MCP server acts only on the Ramly account you connect. Models and results live in that account, and you can view them on ramly.io.
Disconnect any time under Account → Connected AI apps.
See the privacy policy and terms.

Support

Open an issue in this repository or visit ramly.io.