J'ai testé Hermès 2.0 avec Claude Code : Le résultat est fou !

J'ai testé Hermès 2.0 avec Claude Code : Le résultat est fou !

I tested Hermès 2.0 with Claude Code: The result is crazy!

🎙 Parlons IA 👥 17K 📅 June 26, 2026 ⏱ 60 min 👁 20K 📄 tutorial 🧭 2026-09-08
Available in: English (current) Français

Keywords

Hermes AgentClaude CodeAI agentsMCPlocal AI

Summary

The video is a comprehensive tutorial on setting up and configuring Hermes 2.0, an AI agent framework, to enhance its capabilities by integrating Claude Code. The creator demonstrates how to configure the system prompt (Sys.md), connect external tools via MCP (e.g., Airtable, Notion), and create automated workflows. He emphasizes that default Hermes is a ‘shell’ and requires deep configuration to be effective. The tutorial includes a step-by-step guide to installing Hermes, using GLM-5.2 to generate the system prompt, and setting up OAuth connections. It also covers using different LLM providers (MiniMax, Ollama) and running models locally with LM Studio for free. The video concludes with a demonstration of a complete automation for processing Gmail emails into Airtable, and explains how to use ‘skills’ and ‘cron’ for recurring tasks. Throughout, the creator promotes his paid training, but also offers a free Sys.md file. The content is practical and technical, aimed at users with some technical background.

155 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides valuable, actionable information for configuring Hermes Agent, a relatively niche topic. The argumentation is solid in that it explains the technical reasons why default Hermes is limited (e.g., lack of browser control, empty Sys.md) and demonstrates solutions through live examples. The creator’s claims about the power of combining Hermes with Claude are supported by a practical demonstration of extracting DPE data, which Hermes alone cannot do. However, the argumentation is somewhat one-sided, as it heavily promotes the creator’s paid training and makes unverifiable claims about the ‘most powerful AI’ and ‘1000 times more powerful’ without providing benchmarks or external validation.

Scientific Rigor, Source Quality, Title Accuracy

The video does not cite any external scientific sources or research papers. The only source mentioned is the creator’s own website (parlonsia.teachizy.fr) for his training. The technical accuracy of the steps is plausible, but the lack of references reduces the scientific rigor. The title accurately reflects the content, as the video does test Hermes 2.0 with Claude Code and shows impressive results, though the ‘crazy’ outcome is subjective. The comments are overwhelmingly positive, with viewers praising the pedagogical quality and technical depth, though some express difficulty in finding the promised files.

209 words

Title / Content Match

The title accurately reflects the content: the video tests Hermes 2.0 with Claude Code and shows impressive results, though the 'crazy' outcome is subjective.

Quality & Reliability

6/10

The video provides a detailed technical tutorial with practical steps, but relies heavily on the creator's own claims and promotional content for his training. No external scientific sources are cited, and some claims (e.g., '1000 times more powerful') are exaggerated.

Key Moments

Cited Sources

  • Parlons IA - Formations — The creator's website for his AI training, mentioned as the source for the free Sys.md file and paid courses.

Concurring Sources

  • Model Context Protocol (MCP) — The video's use of MCP for connecting tools aligns with the official protocol's purpose.
  • Claude Code — The video demonstrates using Claude Code, which is Anthropic's official coding agent.

Contribution & Novelties

The video offers a practical, hands-on approach to configuring Hermes Agent, which is often presented as a plug-and-play tool. The main novelty is the integration of Claude Code within Hermes to overcome its limitations, and the detailed explanation of the system prompt and MCP configuration. It also provides a free resource (Sys.md) and demonstrates a real-world automation.

Pour aller plus loin :

  • Model Context Protocol (MCP) — The official protocol for connecting AI models to external tools, central to the video’s configuration steps.
  • Claude Code — Anthropic’s coding agent, used in the video to enhance Hermes’ capabilities.
  • LM Studio — The local model runner used in the video to run models for free.
  • Ollama — Another local model runner mentioned for privacy-focused use.
  • Airtable — The CRM used in the automation example, connected via MCP.

135 words

Radar Profile

The radar profile shows high scores in information quantity and technical level, indicating a content-rich tutorial. However, the lower scores in information quality and global reliability reflect the lack of external sources and the promotional nature of the video.

Reliability 5/10

💬 Très positif. Sur les 30 commentaires analysés, les spectateurs expriment une gratitude et une admiration marquées pour la qualité pédagogique et la clarté des explications, certains allant jusqu'à qualifier le contenu de 'pépite' et de 'fossé abyssal' avec d'autres vidéos.