Over the past few weeks, I’ve been building Claude Full Stack 2.0 — a comprehensive collection of production-grade Claude Skills and workflows.
What started as experiments has grown into a Claude Code plugin with 80+ structured skills that turn AI into a true AI-augmented software engineering operating system.
Designed for:
- Founders
- Developers
- Indie hackers
- AI enthusiasts
The goal is to move beyond generic prompting and deliver consistent, high-quality support across the entire product lifecycle:
Idea → Architecture → Development → Production
Key Highlights:
- 80+ skills covering architecture decision domains (tech-agnostic) + ecosystem-specific implementations
- Strong emphasis on DevOps, SRE, observability, security, and production readiness
-
End-to-end workflows including:
- Idea-to-Production Full Stack (Spring Boot + React + Kubernetes)
- Idea-to-Production Flutter (mobile)
- Production Readiness Review
- Cloud Foundation on AWS
Clean architecture with standards, patterns, and quality gates
Now available as an installable Claude Code plugin
Everything is open source under the MIT license. Feedback, contributions, and new skills are very welcome!
Repository: claude-full-stack-2.0
Top comments (2)
80+ skills is a real commitment - I'm running ~60 in my indie content/publishing harness and the management overhead becomes non-trivial past a threshold. The symlink approach (one canonical source, symlinks into project repos) was the only thing that kept it tractable for me; otherwise you end up with three drifted copies of the same skill within a month. Curious how you handle versioning when a skill changes - do you bump anything visible to the user, or just edit in place?
Yeah, it's a full fledged software engineering skills. I haven't have to make changes in the skills yet, but I think I'll update the version if I need to do major changes. I've done lots of planning/research while creating a skills so hopefully I wouldn't need to do so.