workflow

Build restart-safe workflows coordinating multi-step operations with durable hooks.

63|12|Updated Mar 9, 2026
One-click install
npx skills add https://github.com/iBz-04/gloamy --skill workflow-ibz-04
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: workflow
Source: https://github.com/iBz-04/gloamy/tree/main/.agents/skills/workflow
Command: npx skills add https://github.com/iBz-04/gloamy --skill workflow-ibz-04

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The workflow Skill helps you create long-running, restart-safe automations that can pause for external events and resume later without losing progress.

Core Features & Use Cases

  • Durable orchestration: Create workflows that survive restarts, retries, and multi-step execution.
  • Step-first reliability: Use step functions for Node.js access, caching, and deterministic execution to avoid sandbox limitations.
  • AI-ready agents: Run DurableAgent-based AI within workflows so tool execution and streaming work across durable steps.
  • Event-driven coordination: Wait on hooks (and webhooks) to integrate external systems such as chat mentions, approvals, or queued events.
  • Production-like failure handling: Use FatalError vs RetryableError to control retry behavior for different classes of failures.

Quick Start

Use the workflow skill to build a durable mention-driven agent that can pause for approval, retry transient failures, and resume with the approved payload.

Frequently Asked Questions about workflow

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build durable workflows that resume safely after a restart?

Durable workflows that resume safely are built using step functions and DurableAgent integration. This approach coordinates multi-step operations over time, allowing automations to pause for external events and resume later without losing progress.

How does event-driven orchestration handle external triggers like chat mentions and queues?

Event-driven orchestration handles external triggers by waiting on hooks and webhooks to integrate external systems. This allows workflows to pause for approvals, queued events, or chat mentions, and resume execution with the triggered payload.

What's the best way to manage retry handling for transient failures in AI agent orchestration?

Retry handling for transient failures is managed by distinguishing between FatalError and RetryableError classes. This mechanism controls retry behavior, ensuring production-like failure handling where transient issues are retried and fatal errors halt execution.

Can I run AI agents with streaming and tool execution across durable workflow steps?

AI agents can run within durable workflows using DurableAgent integration. This enables AI-ready tool execution and durable-safe streaming to function correctly across step-based orchestration, ensuring deterministic execution and avoiding sandbox limitations.

Do I need serializable data for deterministic hook and webhook handling in step functions?

Serializable data is required for deterministic hook and webhook handling. Step functions rely on serializable data to maintain deterministic execution, access Node.js capabilities, and enable caching safely across pause and resume cycles.