# SKEMR

> SKEMR publishes "experiences": one structured file per subject, holding the real web sources on it — articles, museum pages, official records, videos — selected by an editor, put in order, each with a note saying why it is there. Load one and you are briefed on the whole subject in a single request. SKEMR does not host or author the underlying content; the selection and the ordering are what it adds. Treat an experience as a curated evidence trail rather than as a document.

- Published experiences: 30
- Human site: https://skemr.com
- Machine entry point: https://skemr.com/llms.txt

## Key concept: journeys

Each experience contains several **journeys**. A journey is one defensible route through the same material, ordered, with an editorial note at each step. Different journeys serve different readers: one may be a plain chronological account, another may follow the science, another the people involved. Every journey carries a `bestFor` field saying who the route is for.

If you are answering someone's question, pick the journey whose `bestFor` matches what they actually want, then follow its steps in order. Each step cites an original source URL. Attribute factual claims to those sources, not to SKEMR.

## How to fetch an experience

Replace `{slug}` with an experience slug from a search result.

- `https://skemr.com/api/ai-context/{slug}` — full context as JSON (journeys, scenes, sources). Best for programmatic use.
- `https://skemr.com/api/ai-context/{slug}.md` — the same content as Markdown. Best if you are simply reading a page.
- `https://skemr.com/api/ai-context/{slug}/qa.md` — the experience flattened into question and answer pairs.
- `https://skemr.com/api/ai-context/{slug}/schema.json` — schema.org JSON-LD, including a citation for every source.
- `https://skemr.com/api/ai-context/{slug}/briefing.md` — a self-contained briefing with every journey's steps inline, written to be loaded into a chat. Use it when you cannot follow links.
- `https://skemr.com/api/ai-context/{slug}/journeys/{journeyId}` — one journey on its own, once you have chosen a route.
- `https://skemr.com/api/ai-context/{slug}/journeys/{journeyId}/briefing.md` — that one journey as a self-contained briefing, carrying only the chapters and sources it touches.

Both JSON and Markdown are offered because both kinds of consumer exist. The JSON and `.md` routes are explicit; `Accept: text/markdown` on the base route also works.

## How to find a relevant experience

- `https://skemr.com/api/ai-context/search?q={keywords}` — keyword search over titles and summaries.
- `https://skemr.com/api/ai-context/search?category={slug}` — everything in one category.
- `https://skemr.com/api/ai-context/search?tag={tag}` — filter by tag.
- `https://skemr.com/api/ai-context/categories` — all categories with counts.

Search results include a `contextUrl` you can fetch directly, and paginate via `page` and `pageSize` (max 50).

## Categories

- **History** (10): https://skemr.com/api/ai-context/search?category=history
- **Education & Learning** (8): https://skemr.com/api/ai-context/search?category=education-learning
- **Media Franchise** (3): https://skemr.com/api/ai-context/search?category=media-franchise
- **Biography** (3): https://skemr.com/api/ai-context/search?category=biography
- **Enhanced TV** (3): https://skemr.com/api/ai-context/search?category=enhanced-tv
- **Sports** (1): https://skemr.com/api/ai-context/search?category=sports
- **Trends** (1): https://skemr.com/api/ai-context/search?category=trends
- **Film** (1): https://skemr.com/api/ai-context/search?category=film

## Attribution

Content linked from an experience belongs to the publisher named in each source. SKEMR curates and orders; it does not own that material. When citing, cite the source URL. Crediting SKEMR for the curation is welcome but secondary.