Back to the catalogue

Building agents

Agentic System Card (Document your AI agent)

You set up an agent and it works. What you haven't created yet is the document that explains what it does, what it has access to, who is responsible if it makes a mistake, and how it is turned off. This tool builds that document with you, section by section, following the official IAméricas and Adigital template. There are thirteen sections that go from the purpose of the system to the complete inventory of components. The objective is in the template itself. If there is an incident or an audit tomorrow, any reasonable person has to be able to piece together what the system was, what it could do, what controls it had, and who took responsibility. It adapts to what you already know. If you upload a technical document, the skill reads it and fills in what it can before asking you anything. If you hesitate in a field, it explains what it means, why it matters, and breaks the question down into something easier to answer.

iamericas-tarjeta-sistema-agentico
Download skillv1.2

Free download, no sign-up

What problem does it solve?

The skill starts with two questions that decide how everything is organized: (1) is it a single agent, or are there several coordinating?, and (2) what is the name of the system. Then move forward section by section. You can stop whenever you want and start again later. It never asks you for passwords, tokens or real secrets. It also does not recommend brands. If you don't know what tool to use for a part, it helps you define what that part has to accomplish. If you want to compare real options, it refers you to the Supplier Evaluator.

What does it produce?

A Word and a PDF with the report.
  • Identity and purpose of the system, those responsible and change control.
  • Level of autonomy, memory and limits of action.
  • Tools, permissions and action surface, that is, everything that the agent can touch.
  • Internal architecture, multi-agent topology if applicable, base model and knowledge sources.
  • Evaluation and testing, human supervision, security, privacy and operational compliance.
  • Incident and turnaround plan.
  • The inventory of agent components and evidence annexes.
  • An initial summary with a count of pending gaps and which ones fall into critical sections.

When to use it

  • Before putting an agent into production.
  • A large client or tender asks you for governance documentation.
  • You want the knowledge of the system to not just reside in one person's head.
  • After an incident, to record what was there and what changed.
  • You change something material about the agent, such as its autonomy, its permissions or its data, and it is time to upload the version.

What you need first

  • What the agent does and for whom.
  • Who is responsible for it within your company.
  • What systems and what data do you access, and with what permissions.
  • What you can do without asking permission and what you can never do.
  • What AI model does it use below and if they adjusted it in any way.
  • How they tested it before putting it to work.
  • Who reviews their work and what they would do if it fails or has to be shut down.
  • What you don't know is marked as pending definition, visible and grouped at the end, never hidden within an apparently complete document.

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 · 1.3 MB

    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 · 1.3 MB

    Neutral Spanish throughout. It works out which Python interpreter the system has, states which folder the card validator runs from, warns when the Word document dependency is missing, and `--help` is no longer treated as if it were a filename.

  • v1.0

    1 August 2026 · 1.2 MB

    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