tldraw founder Steve Ruiz (@steveruizok) recommended the prompt "find all of the implicit state machines in my app" on Oct 31 (~7,034 likes, ~13k bookmarks, ~1.41M impressions). The thread resurged on Nov 6-7: Shopify CEO Tobi Lutke quoted it with "It works" (~2,016 likes), Garrett Petersen said it triggered a major refactor and a new agent-skill rule (~1,508 likes), and Yahia Bakour said it found far too many bugs (~4,348 likes). This is community discussion and personal experience, not a vendor launch.

Key Takeaways

  • ✓Original post: Steve Ruiz (tldraw founder), 2026-10-31, ~7,034 likes, 13,095 bookmarks, ~1.41M impressions (public metrics at capture)
  • ✓Resurgence: Tobi Lutke quoted it on Nov 7 with "It works", ~2,016 likes and ~130k impressions
  • ✓Garrett Petersen reported a major refactor and added "no implicit state machines" to his agent skills (~1,508 likes)
  • ✓XState author David Khourshid (@DavidKPiano) posted "State machines are good actually" (~486 likes); some developers warn against blindly ripping them out
  • ✓Framing: personal experience and community discussion, no benchmark data; results vary by codebase
🧭

Turn your technical choice into a development budget

Compare 40 dev plans & simulate token costs vs $20/mo subscriptions

🔬

In-Depth Technical Analysis

Core Background & Industry Pain Points

AI coding agents write code faster than ever, but much app state still hides in boolean flags, if-branches and scattered useState calls. These form implicit state machines nobody has drawn: which state combinations are legal and which transitions dead-end often only shows up as production bugs. On Oct 31, tldraw founder Steve Ruiz recommended the prompt "find all of the implicit state machines in my app". The thread resurged from Nov 6 and became a hot X discussion.

Architecture Highlights & Internals

This is not a new tool but a review lens. The agent reads the codebase, folds scattered flags and branches into explicit states and transitions, then lists illegal combinations, missing states and races. Shopify CEO Tobi Lutke quoted the post with "It works". Garrett Petersen said it led to a major refactor and he added "no implicit state machines" to his agent skills.

Authoritative Benchmarks & Measured Scores

There is no benchmark or controlled study. Public engagement at capture only: the original post has ~7,034 likes and 13,095 bookmarks; Tobi's quote ~2,016 likes; Yahia Bakour's "I genuinely hate how many bugs this found" ~4,348 likes. Heat is not proof of effect, and some developers warn that having implicit state machines does not automatically mean critical bugs.

Developer Hands-on Guide

In Claude Code, Codex or Cursor, first ask the agent for an inventory only (states, transitions, illegal combinations, file locations). Review it, then refactor in small steps with tests. For complex flows, consider an explicit state machine library such as XState. Treat this as community experience, not an official best practice.

Action HubReady to adopt this in production?

Benchmark side-by-side against alternatives, or calculate monthly token cost vs subscription break-even.

Up NextX discussion: Stripe CEO Patrick Collison adds an agent-powered "improvements" box to his vibecoded apps and calls it liquid software
Read Now→