template.yaml

# This file is used by `render workflows init` to scaffold new projects.
# It tells the CLI how to display this template and what instructions to show
# after scaffolding. You can safely ignore it when using this example directly.

name: OpenAI Agent
description: AI-powered agent with tool calling and multi-turn conversations

workflowsRoot: .

renderBuildCommand: pip install -r requirements.txt
renderStartCommand: python main.py

nextSteps:
  - label: Enter your project directory
    command: cd {{dir}}
  - label: Start your local task server
    command: render workflows dev -- {{startCommand}}
    hint: This runs your workflow service locally, allowing you to view and run tasks without deploying to Render.
  - label: Set your OpenAI API key
    command: export OPENAI_API_KEY=your-key-here
    hint: Required for the agent to call GPT-4.
  - label: Run a conversation locally (in another terminal)
    command: "render workflows tasks start multi_turn_conversation --local --input='[\"What is the status of order ORD-001?\"]'"
    hint: The agent will use tool calling to look up the order and respond.
  - label: Deploy your workflow service to Render
    hint: |
      To deploy this workflow to Render, first push this code to a remote repository. Then, create a workflow service from your Render Dashboard, connect it to your repository, and provide the following build and start commands:
      • Build command: {{renderBuildCommand}}
      • Start command: {{renderStartCommand}}
      Your OPENAI_API_KEY can be set as an environment variable during service creation.