● Covered by 1 source · 1 reportMedium impact1 neutral
Diagrid Catalyst 2.0 Introduces Durable and Verifiable Execution for AI Agents
Key points
- Catalyst 2.0 adds failure recovery and cryptographic verification for AI agents.
- Supports frameworks like LangGraph, Microsoft Agent Framework, and Google ADK.
- Durable execution allows resuming interrupted runs without re-doing work.
- Cryptographic verification uses Dapr 1.18 for workflow history signing.
Diagrid Catalyst 2.0 Release
Diagrid announced Catalyst 2.0 on July 28, 2026. This release introduces two core features: failure recovery and cryptographic verification for AI agents. These capabilities are integrated with ten different agent frameworks, including LangGraph, Microsoft Agent Framework, Google ADK, Dapr Agents, AWS Strands, OpenAI Agents SDK, Claude Managed Agents, CrewAI, and Pydantic AI.
Durable Execution for AI Agents
Catalyst 2.0 enables durable execution by representing model and tool calls as durable workflow activities. This allows agents to resume interrupted runs from the point of failure, preventing the need to re-execute already completed steps and incur repeated costs for model calls. The system supports deployment in cloud, on-premise, and air-gapped environments, with durable agents built using a Python SDK.
Cryptographic Verification of Workflow History
The new verification model is based on Dapr 1.18's capabilities. It hashes batches of workflow-history events, links each digest to the preceding signature, and signs the result using the Dapr sidecar's SPIFFE-based identity. This process ensures that deleted, reordered, or modified history is detectable when workflow state is loaded. Receivers can validate each signed chunk against the Dapr Sentry trust anchor, allowing verification of histories outside the originating application.
Implementation Details and Considerations
Dapr's verification features are grouped into Workflow History Signing, Workflow History Propagation, and Workflow Attestation, which carry verified execution context across workflow and service boundaries. It is important to note that signing is disabled by default in Dapr 1.18, requiring the WorkflowHistorySigning feature flag to be enabled and mTLS to be active. Enabling signing for a workflow is a one-way decision; existing unsigned history cannot be retroactively signed, and toggling signing for an active workflow will result in a verification error. Unsigned in-flight workflows must complete or be purged before enabling signing more broadly.
Industry Impact
Yaron Schneider, Diagrid co-founder and CTO, highlighted the need for "proof of what happened" when AI agents interact with tools or delegate tasks. This release addresses a critical requirement for accountability and reliability in complex AI agent systems, particularly as their use expands in enterprise environments.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
Primary sources
GitHub aaif/wg-workflows-and-process-integrationReporting from
Diagrid released Catalyst 2.0, adding failure recovery and cryptographic verification capabilities to AI agents built with various frameworks. This update allows AI agents to resume interrupted multi-step operations without re-executing completed work and provides verifiable proof of execution history, addressing issues of reliability and accountability in AI agent operations.