Diagrid Brings Agentic Durable Execution to Goose

Diagrid, developer of the world’s most reliable foundation for mission-critical workflows and autonomous AI agents and the company behind the Distributed Application Runtime (Dapr), announced a new Diagrid Catalyst integration for Goose, an open source project hosted by the Agentic AI Foundation (AAIF), extending agentic durable execution to the open-source general-purpose AI agent, while also adding support for LangChainGo and Mastra. This release extends Catalyst’s support to over 15 agent frameworks for Python, TypeScript, Java and C#.

Also Read: AiThority Interview with Michael Kwok, Vice President, IBM Bob, and Canada Lab Director

With Catalyst, a Goose recipe commits each step — every LLM call, extension tool call and sub-recipe invocation — before the next one begins. If the process or the node dies, the recipe resumes at the last committed step.

Production agents run for minutes, hours or days. Containers are redeployed, nodes fail and API calls time out. Without durable execution, any of those events restarts the run from the beginning; every completed LLM call, tool call and API request executes again, and the organization pays for the same tokens twice. Catalyst commits each step before the next one begins, so when a process dies, the workflow resumes at the exact step that failed.

Diagrid Catalyst adds durable execution underneath Goose recipes and subagents, so a long-running run resumes at the last committed step instead of restarting. Catalyst plugs into Goose through its standard extension mechanism, so no changes to Goose itself are required. Goose gives developers an open, extensible agent runtime that can operate through desktop, CLI and API interfaces, connect to external tools and systems through model context protocol (MCP) extensions and orchestrate work through reusable recipes and subagents.

With Catalyst, a Goose recipe commits each step — every LLM call, extension tool call and sub-recipe invocation — before the next one begins. If the process or the node dies, the recipe resumes at the last committed step. Work already completed is not repeated and is not paid for twice. The result is a clean separation between the agent experience and the durability layer: developers keep Goose’s recipes, sub-recipes, subagents and MCP extensions exactly as they are, and Catalyst supplies the execution durability beneath them.

Added Support Built for Kubernetes

Diagrid has also added integrations with LangChainGo and the TypeScript framework Mastra, bringing durable execution to organizations developing and deploying agents to Kubernetes backends.

Some agent platforms tie durability to a single hosting environment, but Catalyst takes a different approach. It runs on Dapr, the Cloud Native Computing Foundation (CNCF)-graduated open-source project used in production at thousands of organizations, making agent workloads portable across AWS, Azure, Google Cloud and on-premises clusters. Teams avoid vendor lock-in, with no need to rewrite anything when they move.

“Go and TypeScript agents on Kubernetes hit the same failure every long-running process hits: something breaks, and the work starts over,” said Yaron Schneider, co-founder and CTO of Diagrid and chair of the Agentic AI Foundation Workflows Working Group. “These integrations let developers keep the framework they chose and gain a runtime that picks up where the agent stopped. And since Catalyst is built on Dapr, the same durability runs on any Kubernetes environment, in any cloud or on-premises.”

Also Read: Why Coordination, Not Code, Is the Real Constraint on Shipping Software Today

[To share your insights with us, please write to psen@itechseries.com]