AI Web Scraper – Quick-start Cheat Sheet

→ View the full tutorial here

---

0. One-time setup

# Clone & enter the repo
git clone https://github.com/pgflow-dev/ai-web-scraper.git
cd ai-web-scraper

# Copy the environment file example and add your OpenAI key (required by the tasks)
cp supabase/functions/.env.example supabase/functions/.env
# Edit the .env file and add your OpenAI API key
# OPENAI_API_KEY=sk-...

---

1. Boot the local Supabase stack

npx [email protected] start

---

2. Run all database migrations (table + flow)

npx [email protected] migrations up --local

---

3. Serve the Edge Functions (keep this terminal open)

npx [email protected] functions serve

---

4. Start the worker (new terminal)

curl -X POST http://127.0.0.1:54321/functions/v1/analyze_website_worker

The first curl boots the worker; it stays alive and polls for jobs.

---

5. Trigger a job (SQL editor or psql)

select * from pgflow.start_flow(
  flow_slug => 'analyzeWebsite',
  input     => '{"url":"https://supabase.com"}'
);

---

6. Check results

select * from websites;                 -- scraped data
select * from pgflow.runs;              -- run history

That’s it – scrape, summarize, tag, store!