Deprecated

Workflow Composer has been deprecated and is no longer available. To turn diagrams into workflows, use the Dapr skills instead. See Dapr Skills for Claude Code to get started.

Workflow Composer

Workflow Composer is an AI-powered service to convert workflow diagrams into runnable applications using Dapr Workflows.

Launch Composer

Generate code

Draw or upload a diagram and get a full Dapr Workflow app with activities, config and run files.

Run locally

Use the generated multi-app run file to start everything with the Dapr CLI.

Deploy with Catalyst

Provision backing infra and get workflow visualization with Diagrid Catalyst.

---

Generate

Transform your workflow diagrams into production-ready Dapr applications. Draw or upload a diagram, select your language, and download complete, runnable code.

1. 1
   Create a workflow diagram using draw.io, tldraw, Lucidchart, or export to BPMN diagram.
2. 2
   Upload to Workflow Composer and select your language.
3. 3
   Download your generated Dapr Workflow project.

What you get

The generated project includes:

• Complete Dapr Workflow application in your chosen language
• Activity stubs for each task in your diagram
• Dapr configuration files and multi-app run setup
• Build files and README with instructions

Supported languages & features

Go

Java

JavaScript

Python

.NET

Supported Workflow Features

• Workflow and activities
• Sequential and parallel execution
• Split and join
• Conditional branches
• External events
• Timers and reminders
• Continue-as-new loops
• Workflow triggers

Unsupported Workflow Features

• Multiple workflow processes per app
• Child workflows
• Retry policies

Diagram requirements

Workflow Composer understands BPMN notations and simple flowcharts. Supported formats: JPEG, PNG, GIF, WebP (max 1 MB).

Key elements:

• Activity — Tasks to perform (order, pay, submit)
• Flow — Directional connectors between tasks
• Event — Start, End, Timer, or External Message triggers
• Gateway — Decision points (Exclusive, Inclusive, Parallel)

Customize your code

You own the generated code. Review the structure, replace activity placeholders with your business logic, swap Dapr components, and integrate into your CI/CD pipeline.

---

Run

Test and iterate on your workflow locally with the Dapr CLI.

Prerequisites: Docker, Dapr CLI, and your language runtime installed.

1. 1
   Extract the downloaded project
2. 2
   Navigate to the project directory
3. 3
   Run dapr run -f . to start all services

The multi-app run configuration starts your workflow app and any dependencies automatically. Modify activity implementations and re-run to test changes.

---

Deploy

Deploy to Diagrid Catalyst for managed infrastructure, workflow visualization, and production-grade observability.

1. 1
   Sign up for Diagrid Catalyst (free tier available)
2. 2
   Install the Diagrid CLI
3. 3
   Build your project per the README
4. 4
   Run diagrid dev run -f <multi-app-run-file>
5. 5
   View workflow execution and traces in the Catalyst console

Catalyst provisions backing services (state stores, pub/sub) and provides real-time workflow visualization without additional configuration.

Go deeper

Once you have scaffolded a workflow, continue in your language of choice with the Dapr Workflow SDK guides — .NET, Go, Java, JavaScript, and Python. Inspect executions during local development with the Dapr Dev Dashboard, or walk through the end-to-end durable workflow quickstart if you're starting from scratch.

Next steps

• Fill in the generated activities in your workflow SDK.
• Check the shape of your workflow against the workflow patterns.
• Run the project against Catalyst with diagrid dev run, then Deploy to Kubernetes.

---

FAQ

What image formats are supported?

JPEG, PNG, GIF, or WebP are supported for uploaded diagrams (max 1 MB). Ensure diagrams are clear with readable text and distinct elements.

Can I integrate the generated code with my existing applications?

Yes. You own your code. The generated code follows standard Dapr patterns and can be integrated with existing Dapr applications.

Can I customize the generated code?

Yes. After downloading, you should: 1) Review the generated code to understand the workflow structure and activity implementations, 2) Edit activity implementations to add your business logic, 3) Swap Dapr configurations for your existing services or data sources, 4) Run and extend unit tests to cover your logic, 5) Integrate into your SDLC (version control, CI, deployment).

Who owns the generated code?

You do. Users own all code generated by our AI services, with no restrictions on use, modification, or distribution.

Is the generated code production‑ready?

While we strive for high quality, Diagrid disclaims warranties regarding quality, security, or fitness for a particular purpose. Review and test thoroughly before using in production.

How is my data collected and stored?

Diagrid may use certain data to improve its AI systems; we don't use your data for model training. Collection is limited to metadata and usage patterns.

What if I need higher quotas?

Catalyst Cloud has daily limits to prevent abuse. If you need higher quotas, please reach out.

How do I provide feedback?

We value your input. Join the Diagrid community and share feedback: https://diagrid.ws/diagrid-community.