Claude Code, Cowork, Skills : le guide complet pour aller plus loin

Claude Code, Cowork, Skills : le guide complet pour aller plus loin

Claude Code, Cowork, Skills: the complete guide to go further

Title history
  • Claude Code, Cowork, Skills : le guide complet pour aller plus loin — observed on 20261111-2335 UTC (YouTube)

by ThomRichard · View the channel on PSI

Humanities, Social Sciences & Thought AI — Models & Research UYQArtificial intelligence (AI)UYQFGenerative artificial intelligence
🎙 ThomRichard 👥 6K 📅 May 20, 2026 ⏱ 76 min 👁 31K 📄 tutorial 🧭 2026-11-11
Available in: English (current) Français

Keywords

Claude AIMCPCoworkClaude CodeSkillsSecond BrainFramework CRAFT

Summary

This comprehensive tutorial by ThomRichard, lasting 1h17, teaches how to maximize the value of Claude AI. The video introduces the ‘value equation’ (model × prompt × context) as a guiding framework. It covers Claude’s background (Anthropic, Constitutional AI), subscription plans (free, Pro, Max), native limitations (no image/video generation), and interface settings including memory, artifacts, and code execution. The prompt factor is explored through the CRAFT framework (Context, Role, Action, Format, Tone) and pull prompting. The context factor is the core focus, detailing tools like memory, external context, Deep Research, connectors & MCP, Skills, Projects, Cowork, and Claude Code. The video concludes with building a ‘second brain’ using Claude. Throughout, ThomRichard demonstrates practical examples, such as writing a follow-up email, and provides resources in the description, including a free pack of prompts and frameworks.

133 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides substantial value by demystifying Claude’s lesser-known features (Cowork, Claude Code, Skills) and offering a structured approach to productivity gains. The ‘value equation’ is a pedagogical strong point, clearly separating controllable factors. The CRAFT framework is practical and immediately applicable. The argumentation is largely demonstration-based, with live examples that illustrate the tools’ utility. However, some claims (e.g., limits doubled due to partnership with SpaceX/X) lack evidence, and the presenter’s promotional tone for his own coaching services slightly undermines objectivity. Overall, the reasoning is solid for a tutorial context, though not scientifically rigorous.

Scientific Rigor, Source Quality, Title Accuracy

The video cites no external scientific sources, relying on personal experience and official Anthropic features. The description links lead to additional tutorial videos and the presenter’s resources, but no peer-reviewed studies or official documentation are referenced. The title accurately matches the content, and the chapter structure aids navigation. The presenter’s authority stems from his 25-year digital career, but sources are not verifiable within the video itself. The adéquation between title and content is perfect; the video fulfills its promise of covering advanced Claude usage.

193 words

Title / Content Match

The title accurately reflects the content: the video comprehensively covers Claude Code, Cowork, Skills, and more advanced usage, as outlined in the chapters.

Quality & Reliability

6/10

The video is a practical tutorial by a practitioner with 25 years of digital experience, delivering actionable information about Claude's features. It includes some inaccuracies (e.g., claiming Claude limits were doubled due to a SpaceX/X partnership) without citation, but overall provides a structured framework (value equation, CRAFT prompt framework) that is useful. The absence of citations for technical claims and promotional undertones reduce its scientific reliability.

Chapters

Cited Sources

External References

Contribution & Novelties

The video’s original contribution lies in its structured ‘value equation’ (model × prompt × context) and the practical demonstration of advanced features like Cowork, Claude Code, and Skills in a single tutorial. It bridges beginner and advanced usage by providing a repeatable framework (CRAFT) and emphasizing context as a key differentiator. The notion of building a ‘second brain’ with Claude is actionable, with concrete setup steps.

Pour aller plus loin :

  • Model Context Protocol (MCP) — Official standard for connecting AI to tools and data; central to the video’s ‘connectors’ section.
  • Constitutional AI — Learn about the safety training approach that underpins Claude’s design, mentioned in the intro.
  • Claude Code — Official documentation for the command-line agent, covered in section 5.8.
  • Building a Second Brain — Concept by Tiago Forte, referenced without URL; relates to the video’s final section on knowledge management.

142 words

Radar Profile

The radar profile shows high scores in information quantity (8) and technical level (7), reflecting the video's comprehensive coverage and hands-on demonstrations. However, quality (6) and reliability (5) are moderate due to the lack of external citations and promotional elements, making it a practical but not scientifically rigorous resource.

Reliability 5/10