API

Static JSON, rebuilt with the site. No key and no rate limit. Access-Control-Allow-Origin: *. Licensed CC BY 4.0: quote it and use it, with a link back to the page the fact came from. Every response carries license, attribution and generated.

Which endpoints are there?

Endpoint	What it holds
/api/meta.json	Counts, the generated date, the licence and this list
/api/videos.json	All 720 videos: id, title, channel, company, published, views, likes, form, runtime, what Tom did, the company basis, the era, playable, thumbnail, page and markdown URLs
/api/channels.json	Every channel: videos and views in the tracker, subscribers where fetched, who they are
/api/career.json	38 roles: organisation, role, dates, evidence
/api/numbers.json	The quotable numbers, each with its evidence grade and note
/api/playbooks.json	7 playbooks, 49 principles, with page, markdown and animation URLs
/api/lessons.json	The lessons: question, one line answer, page
/api/events.json	Upcoming appearances, past dates dropped
/api/wiki.json	Every page: path, title, updated, summary, URL, markdown URL
/api/entity-map.json	Typed entities and sourced relations

What changes, and when?

Fields are added, never renamed or removed without a new endpoint. Numbers move as the production tracker updates; generated says when.

Is there anything live?

Two things, both on the same origin: /mcp, the MCP server (Streamable HTTP, read only), and /ask/api, Tom Sharman AI, which has daily limits. See for agents.