Import

filepress import crawls a public blog into Markdown. It discovers URLs (sitemap/RSS preferred), extracts articles and pages, scaffolds a sibling content site, and can ask a local Ollama model for a first-pass theme.

Quick start

From the engine repo:

pnpm install
pnpm filepress import --source https://example.com \
  --inspire https://www.catalystforge.com \
  --yes

Dry-run without writing files:

pnpm filepress import --source https://example.com --dry-run --no-llm

--inspire URLs (up to three) bias structure and accent; later URLs tint fonts and color. --no-llm stays deterministic when Ollama is down.

What lands on disk

A content-only sibling folder: filepress.config.ts, posts/, pages/, static/, and a starter theme.css. Point its dependency at "getfilepress": "link:../filepress", then pnpm install and pnpm dev.

A long source bio becomes pages/home.md and homePage: 'home' (post index at /posts). Old article URLs are written to static/_redirects for Cloudflare Pages / Netlify.

After import

Review before you commit:

• posts/ and pages/ for missing or leftover source chrome
• filepress.config.ts (title, url, homePage, nav)
• theme.css as a first draft, not a finished look
• static/_redirects if old article URLs were captured

Then pnpm install and pnpm dev. Import does not deploy. For local taste-tuning, open Genie in filepress dev. Genie never appears in preview or production. See Genie.

CLI notes: docs/SITE_IMPORT_SPEC.md. A walkthrough lives in Writing.