MD-UBAOS™ — Automation Workflow Architect

The problem

Business owners know a process is broken — someone re-types the same data three times, a lead sits in an inbox for two days — but describing it well enough for an automation platform to build it correctly is hard. Most AI automation prompts skip straight to nodes and miss the exception paths, so the workflow breaks the first time something unusual happens.

What it does

• Runs a short interview in plain language (no automation jargon) to understand the actual bottleneck, who's involved, and what tools are already in use.
• Analyzes the process through a Workflow Intelligence Layer — bottleneck, ROI, exception paths — before designing anything.
• Generates a ready-to-import n8n workflow (JSON), or a precise step-by-step build blueprint if the client already runs on Make, Zapier, or another platform.
• Includes error-handling, credential placeholders, and built-in retry/fallback logic — not just the happy path.
• Enforces Quality Gates covering Data-Mapping Fidelity, Resilience, and Security before delivery.

Frameworks & tools

Outputs n8n-compatible JSON workflows by default, with step-by-step blueprints for Make, Zapier, or other automation platforms when needed.

Why this beats prompting it yourself

Standard prompts generate the happy-path workflow and stop there. This skill runs a structured interview and analysis first, then explicitly designs for what happens when a step fails, a field is empty, or an API times out — the failures that break automations in production, not in the demo.