superdurable/dex

Durable Execution (D-EX) framework built on Temporal/Cadence

Go

110

2,074 commits

updated Sep 20, 2026

See the code

README

Dex - Durable Execution(D-EX)

⚠️ Pre-launch: Dex has not formally launched yet. You can use it for testing, but breaking changes may be introduced before the official launch.

Durable Execution provides programming model that makes an application's execution durable. This includes local state and control flow such as branches and loops, as well as parallel execution and coordination, waiting for timers or external events, error handling, and remote procedure invocations. The application logic is expressed directly in ordinary code, while the platform reliably restores and resumes the execution after failures and restarts.

Dex provides such a structural programming model with only a few concepts as durable primitives. You use Dex to write a Flow filled with ordinary code: durable Steps, Attributes, RPCs, and durable conditions using Channels and Timers. Then you run Workers hosting your Flow. The Client calls Dex Server to start and interact with Flow instances. Dex Server dispatches Step and RPC invocation tasks to your Workers.

Unlike replay-based durable execution engines, Dex does not split your logic into deterministic workflow code and separate activities—Worker handlers are ordinary code, and Attribute data lives in a blob store you can sync to databases you already run.

dex-arch3

Learn more: What is Durable Execution? · Why Dex?

AI coding assistants can use the official Dex Developer skill to build, test, and operate Dex applications through Dex's public programming model. Its source is maintained in superdurable/skill-dex-developer.

💬 Community

🎉 Join the SuperDurable Dex community on Slack to ask questions, share feedback, and connect with other Dex users and contributors.

Quick start

See Quick start on docs.superdurable.io.

Contributors

longquanzheng

1,756 commits

lwolczynski

81 commits

samuel27m

54 commits

duoertai

51 commits

superdurable/dex

Durable Execution (D-EX) framework built on Temporal/Cadence

Go

110

2,074 commits

updated Sep 20, 2026

See the code

README

Dex - Durable Execution(D-EX)

⚠️ Pre-launch: Dex has not formally launched yet. You can use it for testing, but breaking changes may be introduced before the official launch.

Durable Execution provides programming model that makes an application's execution durable. This includes local state and control flow such as branches and loops, as well as parallel execution and coordination, waiting for timers or external events, error handling, and remote procedure invocations. The application logic is expressed directly in ordinary code, while the platform reliably restores and resumes the execution after failures and restarts.

Dex provides such a structural programming model with only a few concepts as durable primitives. You use Dex to write a Flow filled with ordinary code: durable Steps, Attributes, RPCs, and durable conditions using Channels and Timers. Then you run Workers hosting your Flow. The Client calls Dex Server to start and interact with Flow instances. Dex Server dispatches Step and RPC invocation tasks to your Workers.

Unlike replay-based durable execution engines, Dex does not split your logic into deterministic workflow code and separate activities—Worker handlers are ordinary code, and Attribute data lives in a blob store you can sync to databases you already run.

dex-arch3

Learn more: What is Durable Execution? · Why Dex?

AI coding assistants can use the official Dex Developer skill to build, test, and operate Dex applications through Dex's public programming model. Its source is maintained in superdurable/skill-dex-developer.

💬 Community

🎉 Join the SuperDurable Dex community on Slack to ask questions, share feedback, and connect with other Dex users and contributors.

Quick start

See Quick start on docs.superdurable.io.

Contributors

longquanzheng

1,756 commits

lwolczynski

81 commits

samuel27m

54 commits

duoertai

51 commits

Languages

Go

42.5%

TypeScript

17.0%

Python

13.8%

Java

13.5%

Rust

9.7%

CSS

2.4%