This time, I’m excited to share something practical: an open source repo with the templates and playbooks I use to drive my AI-coding workflows:
👉 github.com/DamiMartinez/ai-dev-workflow-templates
This is the third article in my AI workflow series (after
My New Development Workflow
and
Adding References & Context Sources).
Why this repo?
When you work with AI coding agents (Cursor, Gemini CLI, etc.), a lot of the success comes from giving them structured instructions. Over time, I built a set of reusable templates and a base AGENTS.md file that make my workflow faster, more consistent, and easier to scale.
Instead of keeping these private, I wanted to open them up so that:
- Developers don’t need to reinvent templates from scratch
- The community can collaborate and contribute new ones
- We all get a stronger starting point for AI-assisted development
What’s inside
The repo is designed as a community-driven collection of templates and task definitions for common dev scenarios.
Currently it includes:
001_task_planning_template.md→ a systematic framework for planning and executing projects with AI assistance002_bug_fix_template.md→ a structured approach to analyzing and resolving bugs003_code_review_template.md→ a comprehensive template for running consistent code reviews004_git_commit_workflow.md→ a workflow for writing clear and conventional git commits- A base
AGENTS.mdfile to standardize how assistants approach tasks
The idea is that whenever you spin up a coding agent, it can read AGENTS.md, pick the right template, and immediately follow a clear process.
How to use it
- Clone the repo
- Integrate the
AGENTS.mdand templates into your coding tool setup - Use them as starting points for features, bug fixes, reviews, or commits
- Contribute back with your own templates so the community grows
Conclusion
This repo turns the ideas from my previous posts into something you can plug into your own workflow today.
If you’d like to collaborate, check out the repo, try the templates, and submit your own contributions. Together, we can build a shared library of AI-workflow templates that saves everyone time and effort.