# dltHub Context

> Discover and build dlt data pipelines. Every source, destination and pipeline
> page carries a ready-to-use prompt you can paste into a coding agent, plus
> setup guidance and example code for that specific connector.

dlt is an open-source Python library for building data pipelines. This site
catalogues data sources (REST APIs and other systems you can read from),
destinations (warehouses and stores you can write to), and pipelines (a
source/destination pairing with worked setup instructions).

## URL grammar

Construct page URLs directly rather than crawling for them:

- Source page:      https://dlthub.com/context/source/{slug}
- Destination page: https://dlthub.com/context/destination/{slug}
- Pipeline page:    https://dlthub.com/context/pipeline/{source-slug}-to-{destination-slug}
- Catalog by letter:https://dlthub.com/context/resources/{a-z}

Slugs are lowercase and hyphenated, derived from the connector name — for
example "Hugging Face Datasets" is `hugging-face-datasets`.

## Enumeration

- Sitemap: https://dlthub.com/context/sitemap.xml

## Page structure

Every connector page is fully server-rendered; no JavaScript execution is
required to read the content. Each contains:

- An `<h1>` naming the connector, followed by a short description.
- Schema.org `SoftwareApplication` JSON-LD in a `<script type="application/ld+json">`.
- A prompt block — the copy-paste instruction for a coding agent, marked up as a
  distinct panel. This is usually the most useful part of the page for an agent.
- Setup guidance and example Python code in fenced code blocks.
- Related pipelines linking the connector to compatible destinations.

## Notes for agents

- Content is generated per connector and updated over time; treat the page as
  the source of truth rather than any cached copy.
- The prompt blocks are written to be pasted verbatim into a coding agent.
- dlt documentation lives at https://dlthub.com/docs — this site is the
  connector catalogue, not the library reference.