Make money doing the work you believe in
Skill Directory — Week 2 of the Spring Into AI Build Sprint
This week's theme: build something designed for AI to use. I didn't build one more skill — I built the place skills get discovered.
What it is
Skill Directory is a small site + API. Landing page lists each Agent Skill (name, description, when to use, repo, demo). For agents: GET /api/skills returns the same list as JSON. So when an agent needs to "find the right skill," it can call the API instead of guessing. Humans get a simple directory; agents get machine-readable discovery.
I included three skills: my own sprint-helper (random build ideas in JSON), Kyle Sebeysten Azure SQL DTU Profiler (SQL Server metrics in the IDE), and Jake Strait Crypto Market skill (live prices, market overview, trends). All follow the Agent Skills format (agentskills.io).
Why it matters
Week 2 was "tooling for AI tools." The Skill Directory is infrastructure — the layer that makes "which skill do I use?" answerable. Build once, agents (and humans) can discover what's available.
Links
Live: mf-spring-into-ai-week2…
Repo: github.com/myles-mf/mf-…
Shout-out to Jake Strait for the Crypto Market Agent Skill — live prices and market data in JSON, plus a clean demo at crypto-agent-skill.verc….
And to Kyle Sebeysten for the SQL Server DB Profiler skill: an agent-invokable tool that moves metrics into the IDE. Usage like "DTUs comin in hot! Check last 3 hours. Show top 3 offenders" returns a synopsis of top N offending queries over the last N hours, invocation counts, CPU waits, and tuning recommendations. He worked with GPT to identify what metrics SQL Server can expose, then had it produce a suite of SQL scripts plus a PowerShell script that maps natural-language input to script params and aggregates the output into a synopsis — so you can act fast when monitoring alarms go off. Source:
More next week.
