Back to the catalogue

Transparency and evidence

Transparency Lite (Can you show how has your AI been built and how it behaves?)

Transparency does not mean publishing your code. It means that your customers know that they are talking to an AI, that you can explain why the system decided what it decided, and that you have the paperwork to prove it when someone asks for it. This tool interviews you and decides on its own what controls apply to you, depending on where your control boundary is. Buying a ready-made application is not the same as connecting to an API, setting up a search engine on your own documents or training a model. An SME that uses a third-party tool is not required to provide training data from the provider. This registers as a dependency on that supplier, not as a fault on your part. And it measures two things separately, what you declare you do and what you can support with a document.

iamericas-transparencia-lite
Download skillv1.2

Free download, no sign-up

What problem does it solve?

It starts with a triage of a few questions at a time: what you use the AI ​​for, who it affects, whether you bought or set it up, what provider is underneath, what countries you operate in, whether you process personal or sensitive data, whether there are minors involved, and whether the system decides something that affects a person. Answering that you don't know is a valid answer and is thus recorded. It never makes it favorable by inference or presents it as a proven breach. Nor does it accept a statement as evidence. Saying they have a policy doesn't count as having one. If you hand him the document and he can read it, then he does mark it as inspected. If something in your answers raises a regulatory alert, mark it for what it is, a review trigger, never a legal opinion. The result is used to improve internally, to commission a more in-depth evaluation or to decide if you are in a position to start formal Adigital certification.

What does it produce?

A report in Markdown, Word and PDF.
  • Two separate scores, that of what you declare and that of what you can support with evidence.
  • The transparency band your system falls into.
  • The controls applicable one by one, with their level and their evidence.
  • The critical flags, each one with the action it touches.
  • The contradictions that appeared between your own answers.
  • Which depends on your provider and not on you.
  • What you still don't know and you have to find out.
  • The next three most valuable steps.

When to use it

  • A client, tender, or partner asks you how your AI works.
  • You are considering the Adigital Algorithmic Transparency Certificate and want to know if you are close.
  • You want to know what evidence you are missing before someone asks for it.
  • You sell in the European Union, you serve minors or your system decides something relevant about people.
  • You generate synthetic content and want to know if you have to report it.

What you need first

  • What do you use AI for and who does it affect?
  • If you bought it ready-made, you connected it via API, set up a search engine on your documents or trained a model.
  • What supplier and what model is below.
  • In which countries do you operate and if you sell in the European Union.
  • If you process personal or sensitive data, and if there are minors involved.
  • If the system decides or influences something that affects a person.
  • Have what you have written on hand. Contract with the provider, privacy notice, the text that the user sees, screenshots or logs. A document that can be read is worth much more than a statement.

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 · 80 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. Tests for both entry paths are added.

  • v1.1

    14 September 2026 · 77 kB

    Neutral Spanish throughout. It works out which Python interpreter the system has before the first command: on Windows there is no `python3`.

  • v1.0

    1 August 2026 · 72 kB

    First version.