Public Page Fact Monitor

PRIVATE Apify Actor source. Not published to the Apify Store. isPublic=false.

Extracts Organization / Product / Offer facts from public HTML (JSON-LD, Open Graph, title) and emits field-level change events against a previous snapshot you supply.

This tree is deployment-qualification packaging (ARX-REVENUE-R4A). Local repo qualification still uses desk.revenue modules, not the vendored cloud copies.

Will not do

• Scrape LinkedIn, Instagram, Facebook, TikTok, or X
• Bypass login, CAPTCHA, or robots.txt
• Invent missing values as 0
• Call Grok / xAI / other paid models
• Publish to Store, set isPublic, or enable PPE from this package

Input

See input_schema.json. Default allow_live_http=false (fixture-only). A bounded TEST_RUN may set it true for a safe public GET.

Output

• page_facts — normalized facts, schema types, provenance hash
• change_event — field-level from/to when previous facts differ
• access_rejection — robots / login / CAPTCHA / ToS-blocked host / private
• run_summary — counts; model_cost=0

Run locally (repo)

python -m desk.revenue.r4_run
python -m desk.qualify_revenue_r4

Cloud entry is src/main.py (Apify SDK). Local main.py stays argparse-only.

execute_publish and create_apify_account remain hard-refused in repo code. This package does not call them.