franklioxygen

franklioxygen / agent-workflows

Public

Engineering workflows for AI coding agents or flesh engineers.

ai-agents ai-workflows development-guidelines prompt-engineering prompt-workflow
16
3
85% credibility
Found May 27, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Agent Workflows is a library of structured engineering processes designed for AI coding assistants. It provides seven reusable workflows covering common software development tasks: starting new projects, developing features, fixing bugs, reviewing code, handling production incidents, refactoring code, and cleaning up technical debt. The library also includes shared safety rules, preflight checks, and validation conventions that keep AI assistants on track. Additionally, it bundles nine specialized skills that help AI agents select the right workflow, audit the library itself, prepare releases, review code for security or performance issues, plan migrations, design test strategies, and maintain documentation. The project is aimed at developers who want their AI coding helpers to follow consistent, professional engineering practices rather than producing ad-hoc results.

How It Works

1
💡 You want your AI assistant to work more reliably

You've been using AI coding helpers and noticed they sometimes miss important steps or produce inconsistent results.

2
📋 You discover a library of proven engineering processes

This collection gives your AI helper a structured playbook for common tasks like starting projects, fixing bugs, or reviewing code.

3
🎯 You match your task to the right workflow

Seven workflows cover the main scenarios: new projects, features, bug fixes, code reviews, incidents, refactoring, and cleanup work.

4
You let your AI assistant follow the workflow
🔍
Automated mode

Your AI assistant picks the workflow automatically and runs it end-to-end

📖
Guided mode

You follow along step-by-step, using the workflow as a checklist

5
✅ Your work gets validated automatically

Built-in checks confirm your changes work correctly before moving to the next phase.

🎉 Your task is complete with professional structure

Whether it's a new feature, a bug fix, or a code review, your work follows engineering best practices consistently.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is agent-workflows?

agent-workflows is a library of structured engineering processes for AI coding agents. It provides reusable workflows for common tasks like project initialization, bug fixes, feature development, code review, incident debugging, refactoring, and tech debt cleanup. The library is written in Python and bundles Codex skills that help agents choose and execute the right process for a given task. Workflow-specific guidance is separated from shared safety rules and preflight checks, making the documentation easier to maintain and reuse.

Why is it gaining traction?

The project stands out by treating AI agents as junior engineers who need process guidance, not just raw capability. Instead of hoping an agent figures out the right approach, you give it a structured workflow to follow. The bundled scanning scripts for security, performance, and migration signals give agents concrete things to look for during reviews. This turns ad-hoc AI assistance into something resembling a disciplined engineering practice.

Who should use this?

Teams integrating Codex or similar coding agents into their development pipeline will find the most value. It is particularly useful for engineering managers who want consistent behavior from AI-assisted work, or for teams that have seen inconsistent results from unguided agent interactions. Organizations with complex codebases that need structured approaches to refactoring, migrations, or incident response will benefit most.

Verdict

At 16 stars with a credibility score of 0.85%, this is an early-stage project that shows promise but lacks the community validation of more mature tools. The documentation is solid and the workflow separation is well thought out, but test coverage and production hardening remain unclear. Teams should evaluate it for their specific agent integration needs rather than adopting it as a primary workflow engine.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.