Why AI-Generated Code Breaks and Needs Guardrails

Wes and Scott talk about making AI coding more reliable using deterministic tools like fallow, knip, ESLint, StyleLint, and Sentry. They cover code quality analysis, linting strategies, headless browsers, task workflows, and how to enforce better patterns so AI stops guessing and starts producing maintainable, predictable code.

Show Notes

• 00:00 Welcome to Syntax!
  • Losing two clients in one week
• 04:49 Code quality tools
  • jscpd.dev
  • knip.dev
  • fallow.tools
  • wallace
• 14:11 Finding and using components
  • Storybook AI
• 17:28 Brought to you by Sentry.io
• 17:42 Finding bugs
  • Sentry CLI
  • Spotlight
• 19:55 Formatting and linting
  • Vite+
  • ESLint
  • StyleLint
  • clint
• 25:41 Headless browsers
  • agent-browser
  • chrome-devtools-mcp
  • Lightpanda
• 32:11 Tasks and todos
  • dex
  • beads
• 33:32 Docs
  • Context7
• 34:22 TanStack Code Mode
• 36:01 Getting AI to use these tools
• 38:18 Sick Picks + Shameless Plugs

Sick Picks

• Scott: Leaf Two Razor
• Wes: Puzzles

Shameless Plugs

• Phases.fm Podcast

Hit us up on Socials!

Syntax: X Instagram Tiktok LinkedIn Threads

Wes: X Instagram Tiktok LinkedIn Threads

Scott: X Instagram Tiktok LinkedIn Threads

Randy: X Instagram YouTube Threads