Connect Flowise to the Enterprise MCP Server - Teradata Enterprise MCP

Teradata® Enterprise MCP User Guide
Product
Teradata Enterprise MCP
Release Number
2.2.0
Published
June 2026
ft:locale
en-US
ft:lastEdition
2026-06-10
dita:mapPath
tjf1773879711738.ditamap
dita:ditavalPath
ayr1485454803741.ditaval
dita:id
tjf1773879711738
Prerequisite
  • Flowise instance running and accessible
  • Teradata Enterprise MCP server endpoint URL
  • LLM API key (such as, OpenAI, Azure OpenAI, or Anthropic)
  • Basic understanding of Flowise AgentFlows
  • Using Teradata’s customized Flowise build that includes pre integrated MCP Server components (tdflowise image).
  • Using AgentFlows v2. If you use a different version, refer to the Flowise documentation.
Flowise is a low code platform for building AI workflows using visual graphs. This procedure explains how to connect Flowise to the Teradata Enterprise MCP Server so that an agent can use Enterprise MCP Tools and interact with Teradata databases.
  1. Use a browser to go to Flowise.
    http://<FlowiseHost>:3000
  2. Sign in and open the Flowise dashboard.
  3. Create a AgentFlow.
    1. In the navigation panel, select Agentflows.
    2. Select Add new.
    3. Name your workflow (for example, Teradata Query Agent).
      The AgentFlow canvas opens.

    4. Build a visual graph (agent workflow)

    The workflow structure is: Start → Agent → End

  4. Add Start node.
    1. Select and search for Start in the nodes palette.
    2. Drag the Start node onto the canvas.
  5. Add an Agent node from the nodes palette to the canvas.
  6. Connect the Start node output to the Agent node input.

    7. Configure the LLM

  7. Configure the Model parameters.
  8. In the nodes palette, search for your LLM provider, such as ChatOpenAI or ChatAnthropic.
  9. Drag the model node onto the canvas.
  10. Connect the model node to the Chat Model input of the Agent node.
  11. In the model node configuration,
    1. Select or create credentials (or API key) for the provider
    2. Set the model name, such as gpt-4o or claude-3-5-sonnet.
    3. Set Temperature to 0.7 for balanced responses.

    12. Configure Teradata Enterprise MCP Server connection

  12. In the nodes palette, search for Custom MCP and drag it onto the canvas.
  13. Connect the output of the Custom MCP node to the Tools input of your Agent node.
  14. Configure the Custom MCP node.
    1. Transport Type: Select SSE (for URL-based endpoints).
    2. Configuration: Enter the following JSON in the config field.
      {
"url": "http://mcp-server-endpoint/mcp/",
"headers": {
    "Authorization": "Bearer your-token"
}
}
      • Replace mcp-server-endpoint with your actual MCP server endpoint.
      • Replace your-token with a valid authorization bearer token to access the server.
  15. Select Refresh Available Actions.
    The system retrieves the available MCP tools.
  16. Select the MCP tools that the agent is allowed to use.

    17. Configure the System Prompt

  17. Select the Agent/Worker node.
  18. In the System Prompt field, define the agent's behavior and constraints.
  19. Use a prompt template appropriate guiding the agent's behavior for querying Teradata through MCP tools. See Prompts and Prompt Templates.

    20. Save and Test the AgentFlow

  20. Open the chat panel from the lower-right corner of Flowise.
  21. Test the connection by entering a test prompt, such as "What tables are available in my database source?"