Arrêter de Prompter Claude. Utilisez la Méthode de son Créateur

Arrêter de Prompter Claude. Utilisez la Méthode de son Créateur

Stop Prompting Claude. Use Its Creator's Method

🎙 Yassine Sdiri 👥 280K 📅 September 20, 2026 ⏱ 17 min 👁 23 📄 tutorial 🧭 2026-09-20
Available in: English (current) Français

Keywords

Claudeagentic AIskillstask automationloops

Summary

This video by Yassine Sdiri presents three practical reflexes for using Claude, as practiced by its creator Boris Cherny. The story begins with a tweet about a man using Claude Code to grow tomatoes, which inspired the development of Claude Cowork, a version for non-coders. The first reflex is to write principles once in global instructions or skills, avoiding repetition in every conversation. The second is to ask Claude to build a tool rather than repeating the same task, saving time in the long run. The third, less commonly used, is to launch autonomous loops that work continuously without intervention. The video also discusses how Boris’s own workflow evolved from manual prompting to running hundreds of agents, and concludes with reflections on taste and human judgment, acknowledging that some tasks, like translating poetry, require human discernment. Throughout, Sdiri offers a free starter kit and promotes his AI training school, making the content both educational and promotional.

156 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides valuable, transferable insights for users of AI tools like Claude. The argumentation is logical and builds on concrete anecdotes, such as the tomato farming example, to illustrate the need for accessible AI. The three reflexes are presented with clear explanations and practical implications, making the advice easy to grasp and apply. However, the reasoning relies heavily on the authority of Boris Cherny without independent verification, and the promotional segments for Sdiri’s community and school may reduce perceived objectivity. Still, the core content is solid and genuinely useful for improving interaction with AI agents.

Scientific Rigor, Source Quality, Title Accuracy

The video cites Boris Cherny as its primary source, but does not provide direct references to talks or papers. The information is delivered as a summary of Cherny’s practices, with no formal citations. The title is well-aligned with the content, as the central message is indeed to stop conventional prompting and adopt a more systematic approach. The description contains only promotional links to Sdiri’s services, which are not scientific sources. Overall, the rigor is moderate: the content is internally consistent and based on recognized expertise, but lacks verification or external documentation.

202 words

Title / Content Match

The title accurately reflects the content, which focuses on moving beyond simple prompting to using Claude with principles, tools, and loops, directly addressing the creator's methods.

Quality & Reliability

8/10

The video offers practical, well-structured advice based on the practices of Boris Cherny, a credible figure in AI development. It relies on anecdotal evidence and personal endorsement rather than rigorous citations, but the information is coherent and actionable for non-technical users.

Key Moments

Cited Sources

Contribution & Novelties

The video synthesizes Boris Cherny’s personal workflow into three actionable reflexes for non-technical users, filling a gap between complex technical AI usage and simple chatting. It emphasizes moving from single-command prompting to building reusable structures (skills, tools) and autonomous loops, which represents a shift toward agentic AI. While the ideas are not entirely novel, the clear presentation and practical examples make them accessible to a broader audience.

Pour aller plus loin :

  • Prompt engineering — Understand the fundamentals of prompt design to better apply the methods discussed.
  • Intelligent agent — The concept of autonomous agents beyond simple prompting, central to the loop-based approach.
  • Anthropic — Background on the company behind Claude and its product philosophy.

115 words

Radar Profile

The profile shows high scores in quality and reliability, with moderate quantity and technical level. This indicates that the video provides well-presented, credible advice without being overly technical, making it suitable for a general audience interested in practical AI applications.

Reliability 7/10