Back to the catalogue

Building agents

El Perito de La Agentería (Can your AI idea be built?)

Someone introduced you to an AI project. It can come from a supplier, a consultant, your nephew or the La Agentería skill. The Expert reads it and analyzes it critically. Score seven dimensions and calculate a score out of 100. Three of them have double weight because they are the ones that sink the projects, who supervises, where the data comes from and how the result is measured. If the proposal does not say who reviews, it is not taken for granted. It marks it as absent and applies a limit, even if the rest of the proposal is good. The Expert does not fix the proposal, but rather scores it and tells you exactly what should be added.

iamericas-agenteria-perito
Download skillv1.2

Free download, no sign-up

What problem does it solve?

You send him the proposal in the format you have. A result of the La Agentería skill, a PDF from the supplier, an email, some loose notes or pasted text. You don't have to adapt it to anything. If what you give him is so brief that it does not even allow for an indicative evaluation, he tells you and asks you for the minimum instead of inventing a proposal to be able to rate it. When it applies a cap it explains why. For example, the fact that your proposal has a low score in metrics because there were no figures explains the result much better than the final grade alone. A low grade does not always mean that the idea is worthless, but rather that specificities are lacking. And a high grade does not guarantee that the agent works, it measures the quality of the specification, not that of the built system.

What does it produce?

A report in Word and PDF with the score out of 100 and the verdict.
  • The verdict ranges from Not Deployable as-is to Ready to Deploy, through In Exploration Phase, Can Pilot with Major Improvements, and Ready with Minor Tweaks.
  • The seven dimensions scored with their weight.
  • Real strengths of the proposal.
  • Critical leaks ordered by severity.
  • Mandatory improvements and recommended improvements, separate.
  • Pending questions that you can send as is to the person who made the proposal.
  • The level of real agency of the system, which sometimes does not coincide with what the proposal declares.

When to use it

  • Before signing with an AI vendor.
  • Before investing your team's time in a pilot.
  • The proposal sounds good but you don't know what to ask.
  • You want to review your own project before showing it to a client, your board, or an investor.
  • You received several proposals and you want to compare them with the same criteria.

What you need first

  • The proposal, in whatever format.
  • At a minimum, what the agent does, who uses it and what data it works on.
  • If the proposal does not say who supervises or does not put figures in the metrics, do not complete it yourself so that you get a better grade. That is exactly what the report has to point out.
  • It does not need access to your systems, nor credentials, nor personal data of your clients.
  • If you have doubts, use the Agentry skill first.

Where it works and how to install it

Pick your platform and follow the steps. Where the tab is labelled “Native” the file installs as-is; ChatGPT and Gemini do not install it yet, so there you paste the content of SKILL.md into the chat.

Method A — Upload the file (easiest)

  1. 1Download the file with the button above. No need to unpack it.
  2. 2Open Claude — the desktop app on Windows or Mac, or claude.ai — and go to Customize → Skills.
  3. 3Click “Add” and upload the file as it is. If the form rejects the extension, rename .skill to .zip and upload again: same file.
  4. 4The skill shows up in your capabilities list, ready to switch on.

It is installed. Turn it on from the capabilities panel, or type @ followed by its name inside any conversation.

Version history

  • v1.2

    21 September 2026 · 481 kB

    Changes how the conversation starts. If the first message brings no context, the skill explains how the process works before asking anything: the phases, that it can be paused and resumed, and what you get at the end; and it notes that the input can be typed or attached as a file. If the context is already there, it confirms in two lines and gets going without that explanation. The description is rewritten to state that this is a guided, step-by-step process.

  • v1.1

    14 September 2026 · 478 kB

    Neutral Spanish throughout. Reference material that was no longer used has been removed, so the download is smaller than before. It works out which Python interpreter the system has and warns when the Word report dependency is missing.

  • v1.0

    1 August 2026 · 501 kB

    First version.

Other skills on the same topic

Available

The Agent Designer (design your own agent)

Have you never created an agent? Do you want to do it safely and responsibly? Turn a task that takes up hours into the complete design of an AI agent, enriching your idea with steps, safety guardrails, etc.

25 downloads
IAméricas
DownloadView skill