Model Context Protocol moderate risk

Difyworkflow

The mcp-difyworkflow-server allows execution of pre-defined Dify workflows via API calls, enabling integration with language model agents. Requires careful API key management.

Connections & Capabilities

Connects To

GitHub

Capabilities

exec

Quickstart

Config

{
  "mcpServers": {
    "mcp-difyworkflow-server": {
      "command": "mcp-difyworkflow-server",
      "args": ["-base-url", "http://localhost/v1"],
      "env": {
        "DIFY_WORKFLOW_NAME": "workflow-translator“,workflow-genImag",
        "DIFY_API_KEYS": "appkey-xxxxxxxxxxxa,appkey-xxxxxxxxxxxb"
      }
    }
  }
}

Exposed MCP Tools (2)

safe

list_workflows

Lists the available Dify workflows that can be executed.

Read-only operation, no side effects.

moderate

execute_workflow

Executes a specified Dify workflow with a given input.

Executes pre-defined workflows, but input sanitization is crucial.

Safety Assessment

This tool is relatively safe if Dify workflows are designed with security in mind and API keys are managed properly. Risks increase if workflows handle sensitive data or interact with external systems without proper validation and authorization.

Workflows are pre-defined, limiting ad-hoc code execution.
Access is controlled via API keys.
No direct filesystem access.
No shell execution.

API keys must be securely managed to prevent unauthorized workflow execution.
Vulnerable to injection if input sanitization is not performed in the Dify workflow.
Potential for data exfiltration depending on the Dify workflow's actions.
If the Dify platform is compromised, this tool is also compromised.