Skip to main content

Quickstart: Multi-Agent Workflow

This quickstart shows how to run a durable, multi-agent workflow using Dapr Agents and Workflows with Catalyst Cloud. The workflow acts as a customer support system that triages support tickets and provides troubleshooting resolutions through two cooperating AI agents.

You will learn how to:

  • Provision a Catalyst project and related resources using the Diagrid CLI.
  • Run multiple AI agents in a durable workflow using Dapr Agents
  • Monitor multi-agent execution in the Catalyst web console

1. Prerequisites

Before you proceed, ensure you have the following prerequisites installed:

2. Log in to Catalyst

Authenticate to Diagrid Catalyst using the following command:

diagrid login

This command opens a new browser window where you log into Catalyst. Once logged in, you'll be shown a confirmation code (matching the code in your terminal) that you need to confirm.

Confirm your user details are correct using the following command:

diagrid whoami

The expected output contains the name of the organization, your user name, and the Catalyst API endpoint.

3. Clone Quickstart Code

Clone the quickstart code from GitHub:

git clone https://github.com/diagridio/catalyst-quickstarts.git

Navigate to the quickstart directory:

cd catalyst-quickstarts/multi-agent-workflow/python

4. Configure OpenAI API Key

Add your OpenAI API key to resources/agent-llm-provider.yaml:

metadata:
- name: key
value: "YOUR_OPENAI_API_KEY"

Optionally, to use Anthropic instead of OpenAI, update resources/agent-llm-provider.yaml with:

apiVersion: dapr.io/v1alpha1
kind: Component
metadata:
name: anthropic
spec:
type: conversation.anthropic
metadata:
- name: key
value: "YOUR_ANTHROPIC_API_KEY"
- name: model
value: claude-3-5-sonnet-20240620

5. Install Dependencies

Create a virtual environment:

python -m venv .venv

Activate the virtual environment:

source .venv/bin/activate

Install dependencies:

pip install -r requirements.txt

6. Run with Catalyst Cloud

The diagrid dev run command creates your Catalyst Cloud project (if needed), provisions resources (App IDs, Components, managed state stores for execution/memory/registry, and pubsub), configures environment variables, and sets up the connection between your local environment and Catalyst Cloud.

diagrid dev run -f dev-python-multi-agent-workflow.yaml --project dev-python-multi-agent-workflow --approve
tip

Wait for the following log output to confirm the application is running before proceeding:

== APP - customer-support-system == INFO:     Application startup complete.
== APP - customer-support-system == INFO: Uvicorn running on http://0.0.0.0:8003 (Press CTRL+C to quit)

7. Interact with Multi-Agent Workflow

The customer support workflow orchestrates two agents as child workflows with a decision point based on the triage result:

Test the Multi-Agent Workflow using the terminal or the REST Client extension with the test.rest file.

7.1 Trigger the workflow

From another terminal, trigger the workflow via REST API:

curl -i -X POST http://localhost:8003/workflow/start \
-H "Content-Type: application/json" \
-d '{"customer": "Alice", "issue": "My Dapr system fails to start in production."}'

The workflow will:

  1. Run the Triage Agent (check entitlement and urgency)
  2. Run the Expert Agent (retrieve environment info and generate a resolution)
  3. Return a customer-ready response with the proposed fix

The response will include the triage assessment and a detailed troubleshooting resolution from the expert agent.

7.2 Test non-entitled user

Try a query with a customer who does not have an entitlement to see how the workflow handles rejection by skipping the Expert Agent step:

curl -i -X POST http://localhost:8003/workflow/start \
-H "Content-Type: application/json" \
-d '{"customer": "Bob", "issue": "The Kubernetes Jobs controller keeps crashing on startup, in production!"}'

The response will indicate that the customer is not entitled to support, and the Expert Agent step will be skipped.

8. View in the Catalyst web console

Open the Workflow viewer in the Catalyst Cloud web console and navigate to the Workflows section. Select the workflow instance that corresponds to your multi-agent execution.

The Catalyst workflow visualizer shows the orchestrating workflow calling two agents as child workflows. You can navigate into each child workflow to inspect the details of what each agent is performing, giving you full visibility into the relationship from the orchestrating workflow down to each agent's execution.

workflow viewer

9. Clean Up

Stop the running application by pressing Ctrl+C in the terminal where diagrid dev run is running.

Delete the Catalyst Cloud project to clean up all provisioned resources:

diagrid project delete dev-python-multi-agent-workflow

Summary

In this quickstart, you:

  • Built a multi-agent customer support system with triage and expert agents using Dapr Agents
  • Ran the workflow locally connected to Catalyst Cloud for durable execution and state persistence
  • Observed multi-agent orchestration and workflow visualization in the Catalyst console

Next steps