AI Skills
A hands-on review series. I install each Claude skill or agent tool, test it on my own projects, and give an honest verdict.
Impeccable: The Skill That Finds the AI Slop in Your Frontend
Marketing Skills: 47 Agent Skills That Turn Claude Code Into a Marketing Team
Apify MCP: I Gave Claude 5,000 Web Scrapers for Under $0.50
grill-with-docs: The Skill That Asks Before It Codes
How these reviews work
Every review follows the same process. I install the skill or tool on my own machine and run it against a real project, usually one of my MCP servers or the Kubernetes homelab, not a toy example. I test the features the documentation promises, try to reproduce the problems other users report, and note what breaks along the way.
Costs and timings come from the actual runs, so when a review says a job cost two cents or found a flaw in 63 seconds, that number came from a terminal, not an estimate. Each review ends with a verdict: what the tool is good at, where it fails, and whether I keep it in my own setup. When an update changes a verdict, I test again. New reviews land here first and go out to the newsletter the same day.