Agent Layer

Event Scraper Agent

An agent that scrapes 15+ event sources, deduplicates against existing records, and loads clean data into Supabase for newsletter generation.

The event scraper agent runs before each newsletter cycle and pulls from 15+ configured sources (venue websites, event aggregators, community calendars). Each scraper is idempotent, safe to re-run without creating duplicates. The deduplication layer matches on venue + date + title fuzzy match, handling the reality that the same event appears differently across sources. Clean records land in Supabase with normalized fields: title, date, venue, description, original_event_url (never scraper URLs, those are internal only). The agent handles the data plumbing so the newsletter pipeline agent starts with clean, queryable event data instead of raw HTML.

Where it shows up:

Pre-newsletter data collectionBmoreFamilies pipelineDMVFamilies pipelineEvent database maintenance

Build your Knowledge OS

90 minutes from zero to your first skill chain.