mcp-companion

Qlerify Plugins

Official Qlerify plugins for AI coding assistants — domain model sync, data export, and more. Works with Claude Code, Gemini CLI, and Cursor.

Prerequisites

1. Qlerify account with a workflow created
2. Qlerify MCP server configured with your API token (see setup per tool below)

Installation

Claude Code

Configure MCP server in ~/.claude.json:

{
  "mcpServers": {
    "qlerify": {
      "type": "url",
      "url": "https://mcp.qlerify.com",
      "headers": {
        "x-api-key": "YOUR_API_TOKEN"
      }
    }
  }
}

Install the plugin:

/plugin marketplace add qlerify/qlerify-plugins
/plugin install mcp-companion@qlerify-plugins

After installation, skills are available as /mcp-companion:workflow-creation, /mcp-companion:sync, and /mcp-companion:download.

Gemini CLI

Install each skill using the --path flag:

gemini skills install https://github.com/qlerify/qlerify-plugins.git --path plugins/mcp-companion/skills/workflow-creation
gemini skills install https://github.com/qlerify/qlerify-plugins.git --path plugins/mcp-companion/skills/sync
gemini skills install https://github.com/qlerify/qlerify-plugins.git --path plugins/mcp-companion/skills/download

Configure the Qlerify MCP server in ~/.gemini/settings.json per Gemini CLI docs.

Cursor

1. Open Cursor Settings (Cmd+Shift+J)
2. Go to Rules > Add Rule > Remote Rule (GitHub)
3. Enter: https://github.com/qlerify/qlerify-plugins

Configure the Qlerify MCP server in Cursor's MCP settings.

Get your API token from the Qlerify UI.

Plugins

mcp-companion

Teaches AI agents how to effectively use Qlerify's MCP server. Contains the following skills:

workflow-creation

Guides AI agents through building complete Qlerify workflows from scratch — lanes, groups, domain events, entities, commands, read models, cards, and bounded contexts. Includes a full tool reference and a worked e-commerce example.

Triggers:

• "create a workflow"
• "build a domain model"
• "set up domain events"
• "add commands and read models to workflow"
• Any request involving building a Qlerify workflow or adding structural elements

What it does:

1. Follows an 8-step creation sequence (lanes → groups → events → entities → commands on events → read models on events → bounded contexts)
2. Provides best practices for naming, field modeling, and entity relationships
3. Supports nested fields on commands and read models for related entity references

sync

Syncs your codebase's domain model with Qlerify. Detects entities, commands, and read models in your code and ensures they match your Qlerify workflow.

Triggers:

• "sync domain model"
• "update Qlerify"
• "sync entities"
• After implementing features that change domain objects

What it does:

1. Scans your codebase for domain objects (entities, commands, read models)
2. Compares with Qlerify workflow
3. Creates/updates/deletes to keep them in sync

download

Fast download any Qlerify data directly to files. Uses curl + jq to bypass AI processing, making it ~100x faster than standard MCP tools for large data.

Triggers:

• "save to file"
• "download workflow"
• "export"
• Any request to save Qlerify data locally

What it does:

1. Fetches data directly via shell commands
2. Pipes to file without AI processing
3. ~1 second instead of 3-5 minutes for large workflows

Usage Examples

# Invoke skills directly
/mcp-companion:workflow-creation
/mcp-companion:sync
/mcp-companion:download

# Or just ask naturally - skills trigger automatically
> create a workflow for an e-commerce order process
> sync my domain model with Qlerify
> download the Cart Microservice workflow to workflow.json
> save the swagger spec for my workflow to api.yaml