A practical project for curious people
Ask a useful question of an existing bundle. Check its evidence and limits. Then make a small collection about something you know. Explorer helps you inspect knowledge; your chosen AI uses the material it can actually access.
Try, inspect, adapt
These featured experiences and the complete catalogue share one maintained editorial source. Applications, bundles and teaching fixtures have different limits.
Which recorded official page explains Tax-Free Childcare, and what does the saved evidence not establish?
Submitted competition candidate: repository, data, build and deployment remain frozen. Metadata does not decide eligibility.
Which record identifies the modern Coventry cathedral, and which names help find it?
The official title and reviewed discovery names are different fields. The tiny fixture is separate from the full and synthetic products.
How does the bundle connect MCP and Anthropic, and what does that reference assert?
A bundle about AI is not an AI service. The records retain their review and lifecycle signals.
Can you trace a claim to its source, identify missing qualifications and explain when the evidence is insufficient?
Independent teaching material. Source boundaries and legal interpretations remain uncertain; a learning pass does not establish entitlement or complete evidence.
Browse the complete example catalogue, including specialist and conditional collections.
Learn by making
Start with the short guides above, or continue through the full course. App creation is optional. Each stage repeats the same rhythm: explain, inspect a worked example, do one task, check it, then recall what mattered. You can stop after any checkpoint and return later.
Pick a subject small enough to finish and name who it should help.
Write the questions your bundle and app must answer before collecting data.
Find sources and decide authority, rights, privacy, freshness and gaps.
Give things stable identities and add only relationships you can explain and evidence.
Author a small OKF 0.2 bundle, then use exact checks instead of guesswork.
Use Reader, Search, Links, Graph, Timeline and Inspect to find defects.
Connect your AI and compare its answers with questions you held back.
Predict, run, inspect and modify working code before making your own learning UI.
Your subject, not ours
A useful first bundle is narrow enough to review yourself: roughly 15–60 concepts and 5–15 questions.
Public services, transport, planning, heritage or local history
Music, films, books, games, sport, art or a personal collection
A course topic, scientific field, historical period or reading list
Policies, APIs, guidance, research papers or organisational knowledge
Your own notes or records — kept local unless you have a safe sharing decision
Choose your own stopping point
Consumers can finish with a sourced answer. Domain experts can check a distinction. Knowledge workers can prepare a reusable collection. Students can learn by changing a working example.
“Show me why this idea helps my subject, let me try it on a real example, and give me a check I can trust.”Choose your audience journey
Your AI is a collaborator, not your evidence
Learn identities, sources, dates, rights and uncertainty before asking the AI to make anything.
Hold questions back, require record citations and count unsupported or near-neighbour answers.
Predict, run, investigate and modify a starter before using AI to create your personal interface.