You’re holding it wrong! The double loop model for agentic coding @TestDouble
You’re holding it wrong! The double loop model for agentic coding @TestDouble
Uploaded June 2025 | Updated September 2026, 2 weeks ago
The prescriptive agentic coding prompt approach pigeonholes the model into a path. You're missing out on AI agents' greatest strength—explorability! Joé Dupuis shares his double loop model for agentic coding that allows for flexibility and iteration.

00:00 - Intro
01:05 - First Loop (Vibe coding)
04:21 - Second loop (Code review)
09:44 - Code complete!
10:19 - Recap / TED talk

Read the companion blog: link.testdouble.com/520863
You’re holding it wrong! The double loop model for agentic codingStandard Ruby Update Live StreamHow to ensure product vibes align with real customer needsTaste as the only moatEffective React Testing (Jeremy Fairbank)Secrets of Great Conference Talks - Practical Public Speaking AdviceHow PR Review skills take agentic coding boundaries to the next levelMastering the Macro MachineHow to Program (Justin Searls)Get in the Box: Joé Dupuis demos Claude Code sandbox set upYour bike computer is recording more than you TtinkThe biggest change made possible by agentic coding: Real time collaboration between
Test Double |

You’re holding it wrong! The double loop model for agentic coding

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER