Set up with AI

The fastest way to integrate Ghost Writr is to let your coding agent do it. This site is built for that: it publishes a machine-readable map of the whole docs corpus, so an agent can read everything it needs in one shot.

The llms.txt story

This site serves two files at its root:

• /llms.txt — a curated map of the documentation: the pages that matter and what each covers. Good for an agent to orient quickly.
• /llms-full.txt — the full corpus inlined into a single file: every concept page, every SDK reference, every example. This is what you point an agent at when you want it to actually write the integration.

Both live at the root of developer.ghostwritr.io:

https://developer.ghostwritr.io/llms.txt
https://developer.ghostwritr.io/llms-full.txt

What the agent needs

Two things, and that’s it:

1. Tell it to read /llms-full.txt so it has the keyless model, the article shape, and the per-framework patterns.
2. Give it your site ID (see Your site ID). It’s the only key — there is no API key.

Recipes by framework

Copy the block for your stack, swap in your real site ID, and paste it into Claude Code, Cursor, or any coding agent.

Next.js (App Router)

⚡ Hand this to your AI agent

Paste this into Claude Code, Cursor, or any coding agent — it has everything it needs to wire up the SDK (swap in your real site ID):

Read https://developer.ghostwritr.io/llms-full.txt.
My Ghost Writr site ID is <YOUR_SITE_ID>.
Add a /blog index and /blog/[slug] article pages for my published articles, with SEO metadata in generateMetadata and ISR using the @ghostwritr/next SDK in this Next.js App Router app.
It's keyless — the site ID is the only key, there is no API key.

Astro

⚡ Hand this to your AI agent

Paste this into Claude Code, Cursor, or any coding agent — it has everything it needs to wire up the SDK (swap in your real site ID):

Read https://developer.ghostwritr.io/llms-full.txt.
My Ghost Writr site ID is <YOUR_SITE_ID>.
Add a content collection backed by the Ghost Writr loader, plus a /blog index and /blog/[slug] pages that render the articles with SEO meta and JSON-LD using the @ghostwritr/astro SDK in this Astro app.
It's keyless — the site ID is the only key, there is no API key.

React Router v7

⚡ Hand this to your AI agent

Paste this into Claude Code, Cursor, or any coding agent — it has everything it needs to wire up the SDK (swap in your real site ID):

Read https://developer.ghostwritr.io/llms-full.txt.
My Ghost Writr site ID is <YOUR_SITE_ID>.
Add a blog._index route and a blog.$slug route with server loaders that fetch my published articles, plus SEO meta from articleMeta using the @ghostwritr/react-router SDK in this React Router v7 app.
It's keyless — the site ID is the only key, there is no API key.

Any other framework (feed core)

⚡ Hand this to your AI agent

Paste this into Claude Code, Cursor, or any coding agent — it has everything it needs to wire up the SDK (swap in your real site ID):

Read https://developer.ghostwritr.io/llms-full.txt.
My Ghost Writr site ID is <YOUR_SITE_ID>.
Add a blog index and an article page that fetch my published articles with fetchArticles and fetchArticle and render the Markdown using the @ghostwritr/feed SDK in this server-rendered app.
It's keyless — the site ID is the only key, there is no API key.

Point your editor's agent at the docs

In Claude Code or Cursor, just say: Read https://developer.ghostwritr.io/llms-full.txt, my Ghost Writr site ID is <YOUR_SITE_ID>, wire up the SDK for this app. The agent pulls the full corpus and everything it needs is in there.

Next

• Quickstart — do it by hand instead.
• Keyless reads — the contract the agent will rely on.
• The article shape — the fields it’ll render.