Overview
This MCP server provides a lightweight bridge between MCP clients and Dify workflows. It enables invoking Dify-based workflows by leveraging MCP tooling to call the Dify API using a configured base URL and App SKs. The server can be installed via Smithery or run manually, and supports two configuration modes: environment variables and a config.yaml file. In environment mode you set DIFY_BASE_URL (the Dify API base URL) and DIFY_APP_SKS (a comma-separated list of App Secret Keys, each SK corresponding to a different workflow). In config.yaml you specify dify_base_url and a list of dify_app_sks. A CONFIG_PATH can be supplied to point to the YAML config when running the server. The README provides integration examples for clients using uvx or uv, including Git-based from-git startup. The README also notes updates that simplify environment-variable usage for cloud deployments. This server targets developers, cloud operators, and Dify workflow owners who want to expose multiple workflows to MCP-enabled clients.
Features
Env-based configuration
Configure base URL and SKs using DIFY_BASE_URL and DIFY_APP_SKS for cloud deployments.
Yaml-based configuration
Define dify_base_url and dify_app_sks in a config.yaml file for local or controlled environments.
Supports multiple SKs (workflows)
Each App Secret Key (SK) can map to a different Dify workflow, enabling multiple MCP-accessible workflows.
Cloud-friendly deployment
Designed to work smoothly on cloud platforms via environment-variable configuration.
Smithery or manual installation
Install the server through Smithery or install/run it manually depending on your setup.
uvx client deployment
Client-side integration via uvx with a git-based from-git source to pull the server.
uv client deployment
Client-side integration via uv with a local clone workflow for running the server.
CONFIG_PATH support
Option to specify a custom path to a config.yaml file using the CONFIG_PATH environment variable.
Who Is This For?
- Developers:Integrate and invoke Dify workflows via MCP in client applications.
- Cloud Operators:Deploy MCP server in cloud environments using environment variables for base URL and SKs.
- Dify workflow owners:Expose multiple Dify workflows as MCP endpoints by configuring multiple App SKs.
