Hey everyone! Just shipped Build Skill, a CLI that scaffolds AI agent skill repositories in one command.

I kept running into the same friction when creating skills: every time I wanted to create a new skills repo, I was manually wiring up folder structures, marketplace configs, GitHub Actions workflows, sync scripts the same tedious boilerplate every time. None of it is hard. All of it is annoying. And if you get the structure wrong, your skills just silently don't work; no error, no feedback, just an agent that doesn't pick up what you built.

I caught myself copy-pasting from old repos more than once, and that's usually my signal that something should be automated.

So I built Build Skill:

Source.

If you're not familiar, Agent Skills are the open standard for packaging instructions, scripts, and resources that AI coding agents can discover and use think Claude Code, Cursor, Codex, and dozens of others. It's a simple idea with a lot of power: you write a SKILL.md file, structure a folder around it, and suddenly your agent knows how to do something new.

The concept is great. The setup? Not so much.

The problem I kept running into

Every time I wanted to create a new skills repository, I found myself doing the same tedious dance. Create the folder structure. Write the SKILL.md with the right frontmatter. Set up the .claude-plugin/marketplace.json. Wire up a plugin.json for each skill. Add a GitHub Actions workflow for validation. Write sync scripts to keep the README and marketplace config in sync as skills get added or removed.

With the advent of clawdbots, it's as if we've all lost our inhibitions and "put our lives completely in their hands."

I'm all for delegating work, but not giving them too much personal/sensitive stuff to handle.

As usual, Y Combinator came up with segments that are worth investing:

1. Cursor for Product Managers
2. AI-Native Hedge Funds
3. AI-Native Agencies
4. Stablecoin Financial Services
5. AI for Government
6. Modern Metal Mills
7. AI Guidance for Physical Work 8. Large Spatial Models 9. Infra for Government Fraud Hunters 10. Make LLMs Easy to Train