workflow

Orchestrate multi-agent software development workflows from instructions to PR preparation.

Updated Apr 10, 2026
One-click install
npx skills add https://github.com/erichugy/agentic.skills --skill workflow-erichugy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: workflow
Source: https://github.com/erichugy/agentic.skills/tree/main/workflow
Command: npx skills add https://github.com/erichugy/agentic.skills --skill workflow-erichugy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill provides a repeatable, instructions-first process to coordinate AI-driven code work so teams avoid unreviewed, no-op, or poorly scoped changes and ensure production-quality commits and PRs.

Core Features & Use Cases

  • Instructions-first workflow: create a self-contained instructions document that embeds the relevant source, types, API contracts, and tasks before any code is written.
  • Branching, worktrees & multi-agent execution: standardizes branch naming, creates isolated worktrees per agent, and dispatches worker and reviewer agents with explicit skills.
  • Rigorous review loop and no-op detection: enforces a reviewer checklist, mandates zero-issue reviewer PASS, prevents committing no-op changes, and orchestrates PR creation without merging.
  • Use Case: run this workflow to implement a new feature across parallel worker agents, iterate through reviewer fixes until PASS, then push and open PRs for final human merge.

Quick Start

Create a self-contained instructions document, get Eric's approval, then run the multi-agent plan to implement, review, and prepare PRs.

Frequently Asked Questions about workflow

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

FAQPage Schema
How do I orchestrate multi-agent execution for parallel feature development?

Multi-agent execution orchestrates parallel feature development by dispatching isolated worker and reviewer agents to separate worktrees. This ensures vetted, review-ready code changes without conflicting branch states across your repository.

What is an instructions-first workflow for AI-assisted code changes?

An instructions-first workflow requires a self-contained document embedding source, types, API contracts, and tasks before writing code. This prevents poorly scoped changes by ensuring AI agents operate within strict, predefined boundaries.

How do I prevent committing no-op changes during AI code generation?

To prevent committing no-op changes, the workflow enforces rigorous review loops and no-op detection checks before pushing. This guarantees that only meaningful, verified code modifications proceed to conventional commits and PR creation.

Does this workflow automatically merge PRs after the reviewer agent passes?

No, the workflow orchestrates PR creation without merging. It mandates a zero-issue reviewer PASS and prepares the PR for final human merge, ensuring human oversight over production-ready commits.

How do I prepare production-ready commits using isolated worktrees?

Production-ready commits are prepared by standardizing branch naming and creating isolated worktrees per agent. The system enforces production hardening checks and conventional commits before opening pull requests.

Can I use this multi-agent workflow for bug fixes and code reviews?

Yes, the workflow applies to feature development, bug fixes, and code reviews. It standardizes branching and dispatches reviewer agents with explicit skills to iterate fixes until passing.