skills.rest
m31uk3
m31uk3COMMUNITY
Explore all skills by this author →
GitHub

ai-workflow-engineering

Design AI workflows with a five-phase structure and human-in-the-loop checkpoints.

7|2|Updated Dec 28, 2025
Search Tags:#workflow#engineering#SOP#phase-based planning#LLM uncertainty
Ready to run
npx skills add https://github.com/m31uk3/ai-skills --skill ai-workflow-engineering
💬 Prompt for ChatGPT / Claude Chat▼
Please help me install this Agent Skill.
Skill: ai-workflow-engineering
Source: https://github.com/m31uk3/ai-skills/tree/main/skills/ai--workflow-engineering
Command: npx skills add https://github.com/m31uk3/ai-skills --skill ai-workflow-engineering

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a universal framework for designing reliable AI workflows and SOPs that manage LLM uncertainty through a clear, phased structure and explicit checkpoints.

Core Features & Use Cases

  • 5-Phase Universal Structure: Intake/Investigation, Decomposition/Planning, Iterative Execution, Validation/Review, and Decision Point.
  • Upfront Parameter Capture: Gather domain, goals, and output location to tailor the workflow.
  • Phase-specific Artifacts: Domain analysis, I/O spec, phase checklists, constraints, and validation patterns.
  • Human-in-the-Loop Checkpoints: Explicit decisions at checkpoints to ensure alignment and reduce drift.
  • Domain Adaptability: Applicable to software, strategy, writing, and research workflows.

Quick Start

Define the workflow domain (e.g., "code review" or "meeting facilitation") and the main goal, then supply:

  • workflow_domain
  • primary_goal
  • target_users (optional)
  • output_location (default: "./workflows")