Status: [LIVE] Endpoints found: https://api.lemonsqueezy.com/v1/stores|variants Source: https://github.com/owerryking-beep/rag-scrape # RagScrape — URL → clean Markdown, at the edge One API call turns any public URL into clean, LLM-ready Markdown. Built for RAG pipelines, knowledge bases, and AI agents. ## Live now - **API + storefront:** - **Source:** - Plans: **Free** 50/mo · **Starter** $9/mo 2 000 · **Pro** $19/mo 10 000 · **Unlimited** $49/mo 1 000 000 (Lemon Squeezy checkout; merchant of record — pays out to Kenya. Switched from Stripe 2026-09-21: Stripe live activation requires US-only identity + bank details.) ### The gaps we fill (why this isn't "another scraper") | Gap | RagScrape | Typical tools | |---|---|---| | Sites have no `/llms.txt` | `POST /llms-txt` crawls & generates one | — | | Re-indexing re-pays for unchanged pages | `contentHash` + `ifNoneHash` → `unchanged:true`, zero content tokens | re-scrape & re-embed everything | | LLM context tokens wasted on links/images | `stripLinks` / `stripImages` (fence-aware) | — | | Scrape → chunk → embed = 3 vendors | `embed: true` → chunks with 384-dim vectors in one call (Workers AI) | stitch it yourself | - Free no-signup web tool: - RAG-ready chunking (`POST /scrape` with `chunk: true`) and a same-host docs crawler (`POST /crawl` → pages + `llms.txt`). - **Module 1** — Cloudflare Worker API (`worker/`): Hono + `@mozilla/readability` (on linkedom's DOM) + a custom Markdown converter. KV-based API keys, monthly rate limits, Stripe Checkout + webhooks. - **Module 2** — Node.js CLI (`cli/`): `npx rag-scrape ` with `ora` spinner, free-key registration, and batch mode. - **Module 3** — GitHub Action (`github-action/`): scrape a URL list, write Markdown files, commit & push. - **Module 4** — Lemon Squeezy payments + landing page (`landing/index.html`): $19/mo Pro subscription (10 000 calls) with Tailwind + Alpine. ``` CLI / Action / y