These samples demonstrate how to build and deploy durable agentic workloads using Dapr Agents, and run them with Diagrid Catalyst.
# Create a virtual environment
python3.10 -m venv .venv
# Activate the virtual environment
source .venv/bin/activate # On macOS/Linux
# .venv\Scripts\activate # On Windows
# Install dependencies
pip install -r requirements.txtexport OPENAI_API_KEY=your_api_key_hereSign up for Diagrid Catalyst and install the CLI: https://docs.diagrid.io/catalyst/getting-started/first-api-call
Now login to Catalyst and create a project with all the infrastructure your agent needs - workflows for orchestration, pub/sub for messaging, and key-value storage for state:
diagrid login
diagrid product use catalyst
diagrid project create dapr-agents-catalyst-samples --deploy-managed-pubsub --deploy-managed-kv --enable-managed-workflowThis provisions a complete managed Dapr environment in the cloud. The project overview shows your managed infrastructure components being deployed:
Each example demonstrates different patterns and capabilities of Dapr Agents with Catalyst. Navigate to each folder and follow the specific instructions in their README files.
| Example | Description |
|---|---|
| 01_non-durable-agent | Start here! Simplest non-durable agent example demonstrating conversation memory and Dapr chat client integration |
| 02_durable-agent-headless | Durable headless agent triggered via REST API and PubSub that demonstrates workflow orchestration for backend integrations |
| 03_durable-agent-chat | Interactive chat agent with Chainlit UI that maintains conversation state and full workflow observability |
| 04_agent-orchestration | Workflow orchestration combining Dapr Conversation API with Dapr Agents for sequential task chains |
| 05_customer-support-system | Complete multi-agent system demonstrating complex workflow patterns and agent coordination |
- Try the 01_non-durable-agent for a quick intro.
Do you have any questions? Reach out to us here: https://www.diagrid.io/contact-us
