Start here: practical tools for thoughtful AI work

Version 1.2 · Published October 13, 2026 · Revised October 13, 2026

This collection brings together five current public companion tools for working thoughtfully with AI. The materials are written for this site, with public references and clearly labelled examples. They are editable starting points for your own review and practice.

What is here

• Plan a visual that explains something: a template for choosing an explanation, checking labels and supplying a text alternative
• Check the file people will actually print: a workflow for inspecting exported pages and recording what was checked
• Review a claim against its evidence: a skill card for matching a conclusion to the observations that support it
• Keep a practical learning journal: a guide for turning one experiment into a bounded, reusable lesson
• This guide: how to choose, adapt and review a material from the collection

Each detail page contains the complete material. You can read it in your browser, copy the full text, or download a Markdown file. No installation or account on this site is needed.

Choose a useful starting point

Start with one concrete problem. If a diagram is confusing, use the visual-explanation planner. If a handout is almost finished, use the print-review workflow. If an AI answer makes a strong claim, use the evidence-review card. If you want to retain what you learned, use the journal guide.

A skill in this collection is a human-readable practice card. Downloading it does not configure an agent or grant permissions. Workflows and prompts are proposed methods, not proof of future results.

Try a small version first

1. Read the purpose, audience and evidence status on the resource page.
2. Copy the file into a text editor, notes app or document tool.
3. Replace bracketed fields with your own permitted material.
4. Use a synthetic example or non-sensitive inputs for the first attempt.
5. Complete the review checks and record what you actually observed.
6. Adapt the result to your audience before relying on it.

For example, try the visual planner on an invented four-step book-swap process. Ask another reader to describe the steps without your explanation. Their answer tells you more about the draft than a general impression that it looks polished. This is a proposed exercise, not a reported study.

Read evidence labels carefully

A public reference supplies background for a method; it does not certify the resource. A synthetic example illustrates a pattern; it is not a measured project outcome. A completed checklist records the checks you performed; it does not establish everything the checklist could possibly address.

Keep three questions visible: What did I try? What did I observe? What remains uncertain? Make your conclusion no broader than the evidence supports.

Record a useful revision

Resource and version used:
Purpose and intended audience:
What I changed:
Why I changed it:
What I tried:
What I observed:
What remains untested:
Next revision or check:

The publication and revision dates on each material describe that public file. Versions help you tell which copy you used. A changed date is not a claim that the underlying method has become validated.

Reuse responsibly

Use only information you are permitted to use. Keep private records, identifying learner information and restricted materials out of a public example. Responsible reviewers should approve context-specific use, especially where a decision has important consequences.

These files are editable working resources. This release does not apply a blanket open-source or Creative Commons license. Linked third-party materials retain their own terms; check them before copying text or images. Public references provide background, not endorsement.

Keep learning

The journal pairs approachable explanations with the evidence and limits behind them. Follow the RSS feed to see newly published editions and resources. A small, reviewed practice is a useful beginning; record what changes when you use it.

Revision note

Revised October 13, 2026: added a practical worked check and a clearer next decision. Original publication and evidence limits are preserved.

Find the wider learning library

This guide introduces the five journal companions. The site also contains 30 complete learning artifacts, divided equally among Educators, Agentic Systems and Workspace Development. The companions and learning artifacts serve different-sized tasks; the page and download of one method are not counted as separate artifacts.

Choose the collection by the output you want:

• A lesson, material, discussion, feedback draft or coaching next step: begin with Educators
• An architecture choice, permission map, tool contract or evaluation record: begin with Agentic Systems
• A task brief, source pack, coordinated bundle or repeatable workflow: begin with Workspace Development

If the topic is unfamiliar, first read the plain-language introduction and the worked example. If the topic is familiar, inspect the inputs, failure conditions and checks before adapting the instructions. In either case, keep the evidence limit visible.

A useful stopping point

Your first attempt is complete when you have one output, have performed the checks available to you, and can say what still needs a person or another source. It does not require finishing every related guide. If a missing permission, essential source or consequential decision prevents progress, write down that blocker instead of substituting a guess.

The resource library connects methods by goal. Return there when you need a next step, rather than adding complexity just because another template exists.

Public references

• ChatwithGPT: The Learning Journal