Codex NO ES SOLO PARA PROGRAMAR. Así puedes crear agentes

Codex NO ES SOLO PARA PROGRAMAR. Así puedes crear agentes

🎙 EDteam 👥 1.0M 📅 May 15, 2026 ⏱ 17 min 👁 72K 📄 tutorial 🧭 2026-08-02
Available in: English (current) Français

Keywords

CodexAI agentsautomationOpenAItutorial

Summary

The video, presented by EDteam, explains how to create AI agents using OpenAI’s Codex, emphasizing that agents are autonomous systems that operate without constant human intervention. The presenter, Álvaro, demonstrates two practical examples: setting up an automation to summarize Gmail emails and send them to Slack, and using Chrome control to perform web tasks. He outlines three methods for creating agents with OpenAI: the Agents SDK for developers, Codex for personal use, and Agent Workspace in ChatGPT for team collaboration. The tutorial highlights the importance of connecting tools (complementos) like Gmail, Slack, and Chrome, and configuring triggers and schedules. The presenter also troubleshoots common issues, such as missing permissions or incorrect directory paths, showing that while the process is relatively simple, it requires attention to detail. The video includes a promotional segment for EDteam’s enterprise AI training and mentions related courses. Overall, it provides a beginner-friendly introduction to creating AI agents with Codex, emphasizing practical application over deep technical theory.

160 words

Critical Evaluation

The video offers a valuable, practical introduction to creating AI agents using OpenAI’s Codex, targeting a broad audience interested in automation without extensive programming knowledge. The presenter, Álvaro, demonstrates a clear, step-by-step approach, making the content accessible and actionable. The demonstrations are particularly effective: the first shows how to create an automation that summarizes Gmail emails and posts them to Slack, while the second attempts to control the Chrome browser to search for a product on Mercado Libre. These real-world examples help viewers understand the potential and limitations of AI agents.

However, the video has several weaknesses. The technical depth is limited; it does not explain the underlying architecture of agents, the role of the Agents SDK, or the differences between Codex and ChatGPT’s Agent Workspace in detail. The presenter mentions these options but does not delve into their specific use cases or technical requirements. Additionally, the troubleshooting segment, while honest about the challenges, could be more systematic; the presenter often relies on trial and error rather than providing a structured debugging approach.

The scientific rigor is moderate. The video is based on the presenter’s personal experience and OpenAI’s documentation, but it does not cite external research or compare with other platforms like Claude or Gemini, which are mentioned as future topics. The promotional content, including a mid-roll ad for EDteam’s enterprise services and mentions of their courses, is clearly separated but still interrupts the flow. The presence of these promotions does not significantly detract from the educational value, but it does reduce the perceived objectivity.

The title accurately reflects the content, and the video successfully delivers on its promise to show that Codex is not just for programmers. The adéquation between title and content is strong, and this does not heavily influence the overall rating.

In terms of sources, the video relies on OpenAI’s official tools and documentation, but no specific URLs are provided in the description beyond EDteam’s own courses and social media. This limits the ability to verify claims independently. The presenter’s explanations are generally accurate, but the lack of citations means viewers cannot easily cross-reference the information.

Overall, the video is a useful resource for beginners looking to automate tasks with AI agents. It provides a clear, practical guide with honest insights into the challenges involved. However, for a more rigorous understanding, viewers would need to consult official documentation and additional resources. The video earns a solid rating for its educational value, but it falls short of excellence due to limited technical depth and reliance on anecdotal evidence.

422 words

Title / Content Match

The title accurately reflects the content, which focuses on using Codex beyond programming to create AI agents.

Quality & Reliability

7/10

The video provides a practical, hands-on tutorial on creating AI agents using OpenAI's Codex, with clear demonstrations and troubleshooting. However, it lacks in-depth technical details and relies on anecdotal evidence, and the promotional segments reduce overall rigor.

Chapters

Cited Sources

Concurring Sources

  • OpenAI Agents SDK — Official documentation for building agents with OpenAI's SDK, aligning with the video's mention of the Agents SDK.
  • OpenAI Codex — Official documentation for Codex, confirming its features and capabilities as described in the video.

Contribution & Novelties

The video provides a practical, hands-on guide to creating AI agents using OpenAI’s Codex, emphasizing that Codex is now accessible to non-programmers. It demonstrates two real-world automations (email summarization to Slack and browser control) and explains the three main approaches: Agents SDK, Codex, and ChatGPT’s Agent Workspace. The main novelty is the focus on Codex as a personal agent tool, contrasting with ChatGPT for team use.

Pour aller plus loin :

145 words

Radar Profile

The radar profile shows balanced scores across information quantity, quality, technical level, and reliability, with a slight emphasis on practical application over deep theory. This indicates a well-rounded tutorial that is accessible yet informative, though not highly technical or exhaustive.

Reliability 7/10