Upload workflows/local_business_hunter.md with huggingface_hub

workflows/local_business_hunter.md

@@ -0,0 +1,36 @@
+# Workflow: Local Business Hunter (Ideal)
+
+**Objective**: Automatically discover and research local leads for a specific niche and location with 0 budget.
+
+## Required Inputs
+- `NICHE`: The industry or business type (e.g., "Dentist", "Plumber", "HVAC").
+- `LOCATION`: The city or area to target (e.g., "Austin, TX", "London, UK").
+
+## Steps
+
+1. **Discovery (google_maps_scraper.py)**
+   - Launch a stealth browser using Playwright.
+   - Search Google Maps for `NICHE` in `LOCATION`.
+   - Scroll through results to find up to 50 businesses.
+   - Extract: **Name**, **Website**, **Phone**, **Rating**, and **Review Count**.
+   - Save to `.tmp/raw_leads.json`.
+
+2. **Enrichment (contact_enricher.py)**
+   - Read `.tmp/raw_leads.json`.
+   - For each business with a website:
+     - Visit the website using `httpx` and `BeautifulSoup`.
+     - Scrape the homepage and "Contact/About" pages for:
+       - **Emails** (using regex).
+       - **Facebook URL**.
+       - **Instagram URL**.
+       - **LinkedIn URL**.
+   - Filter out duplicate or junk entries.
+
+3. **Synthesis**
+   - Combine the scraped data from Google Maps with the enriched contact info.
+   - Save the final dataset as `leads_NICHE_LOCATION.csv`.
+
+## Edge Cases
+- **No Website**: If a lead has no website, highlight it as a "High Opportunity" for web design services but skip enrichment.
+- **CAPTCHA**: If blocked, the agent should alert the user to check the IP/proxy or wait.
+- **Dead Links**: Sites that return 404 or connection errors should be marked as "Inactive".