Kids Activities

Run at each heartbeat during reasonable hours (8am–10pm). Pick 3–4 sources from the Sources DB sorted by Last Scraped ascending (stalest first).

Per-source process

1. Fetch the source's Events / Calendar URL via web_fetch
2. If JS-rendered or no events extracted, fall back to web_search — note in Scrape Notes
3. For each event found:
   • Query Events DB by Event Name + Date to check for duplicates before inserting
   • Insert new events with Source relation linked to the source's page ID
4. Update Last Scraped on the source record to today's date

After each batch

1. Prune — trash events where Date < yesterday (non-recurring) or Date > 3 months from today
2. Regenerate the Activities Page (see Workflow 3)

Workflow 2: Adding a New Source

When Emerson provides a new venue, blog, or aggregator:

1. Add to Sources DB — Name, Type, Website, Active=true, Scrape Notes (any crawl quirks)
2. If it's a venue, add a blurb to the Rainy Day Reference page under the right section
3. Confirm back: "Added [Source]. Will scrape at next heartbeat."

Do NOT create a new top-level Notion page for the source.

Workflow 3: Regenerating the Calendar Page

After every scrape + prune cycle, rewrite the content blocks of the Kids Activities page.

⚠️ Critical rules — read before touching any page

• NEVER trash or delete the Activities Page or Rainy Day Page — these are permanent pages that Emerson owns
• NEVER use in_trash: true on a page — only on individual event records in the Events DB
• To "rewrite" a page: fetch its existing child blocks, delete those blocks individually, then append fresh blocks
• The page itself (32619a09-5fb6-81e9-a178-dd3722bb9d4f) must always survive

Block-replace procedure

1. GET /v1/blocks/{page_id}/children — list current child blocks
2. For each block ID returned: DELETE /v1/blocks/{block_id} — remove old content
3. PATCH /v1/blocks/{page_id}/children — append the new blocks

Page structure

🗓 This Weekend — callout block with events for upcoming Sat–Sun
🔄 Recurring Weekly Options — ongoing drop-ins, markets, etc.
🌷 March 2026
  Weekend Mar 28–29: ...
🌸 April 2026
  Weekend Apr 4–5: ...
  ...
☀️ May 2026
  ...

Event line format: 📅 SAT Mar 28: **Event Name** — Venue, HH:MM, cost tier, age. [Link]

The "This Weekend" callout is the fast answer to "what should we do this weekend?" — always regenerate it.

Scrape Priority & Cadence

• Weekly: ROM, Ripley's, Harbourfront, Young People's Theatre, High Park Nature Centre
• Every 2 weeks: AGO, Evergreen Brick Works, Sorauren, Dufferin Grove, Riverdale Farm, Toronto Botanical Garden, Roundhouse, Adventure Alley, CMCP, Casa Loma, Toronto Zoo, Exhibition Place, Downsview Park, Ontario Science Centre, TIFF
• Monthly/seasonal: Canada's Wonderland, Blue Jays/Rogers Centre, Toronto Island/Centreville, Kortright Centre, Spadina Museum

Full source list with URLs and crawl notes: references/sources.md

What NOT to do

• ❌ Never trash or delete the Activities Page or Rainy Day Page — only delete individual child blocks when regenerating
• ❌ Never use in_trash: true on anything except stale event records in the Events DB
• ❌ Never scrape all sources at once — 3–4 per heartbeat, stalest first
• ❌ Never insert duplicate events — query by Event Name + Date first
• ❌ Never leave Last Scraped blank after a successful scrape
• ❌ Never create a top-level Notion page for a source