workflows-mcp-server

Overview

This MCP server provides a powerful workflow orchestration layer for AI agents. It allows agents to discover, understand, and execute multi-step workflows defined in simple YAML files, enabling more structured tool usage for complex tasks. The server includes features like dynamic workflow discovery, global instruction injection, and support for both permanent and temporary workflows.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• workflow_return_list

  Discovers and lists available workflows with filtering options.

  Try:List all available workflowsTry:Show me workflows tagged with 'data-processing'Try:Find workflows in the 'automation' category
• workflow_get_instructions

  Retrieves the complete definition for a specific workflow.

  Try:Get instructions for the customer_report workflowTry:Fetch version 2.1 of the billing workflow
• workflow_create_new

  Creates a new, permanent workflow YAML file.

  Try:Create a new workflow to process monthly reportsTry:Define a workflow for customer onboarding
• workflow_create_temporary

  Creates a temporary workflow that is not listed in the index.

  Try:Create a temporary workflow to fix this bugTry:Define a temporary plan for analyzing this dataset

Comparable tools

Installation

{
  "mcpServers": {
    "workflows-mcp-server": {
      "command": "npx",
      "args": ["workflows-mcp-server"],
      "env": {
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

npm install workflows-mcp-server
npm run build
npm start

FAQ

What file format are workflows defined in?

Workflows are defined in simple YAML files, making them human-readable and easy to create.

Can workflows be updated dynamically?

Yes, the server automatically discovers and indexes all workflow files on startup and watches for changes. Global instructions can also be injected dynamically into all workflows.

Compare workflows-mcp-server with