Add support for syncing to BigQuery

[patrickeasters]

CheckMeIn data (visits, unlocks, certifications) currently lives only in a local SQLite database with no path to cross-referencing it with other org data. By syncing to BigQuery, we can query it alongside our Salesforce data and use LLMs to answer operational questions that span both systems — e.g. correlating member visit history with membership status, or identifying certified members who haven't been in the shop recently.

Summary

• Adds POST /sync/ — a Bearer token-protected endpoint that reads four SQLite tables and uploads them to BigQuery via full table replace (WRITE_TRUNCATE)
• Supports both service account JSON key file and VM-attached IAM identity (Application Default Credentials)
• Response includes row counts per table: {"status": "ok", "rows_uploaded": {"visits": N, "guests": N, "unlocks": N, "certifications": N}}

Tables synced

Table	Key fields
visits	rowid, start, leave, barcode, status
guests	guest_id, displayName, email, firstName, lastName, whereFound, status, newsletter
unlocks	time, location, barcode
certifications	user_id, tool_id, tool_name, certifier_id, date, level

Note: certifications is joined to tools so tool_name is included directly — no separate lookup needed in BigQuery.

Configuration

Add to production.conf:

sync.token           : 'your-secret-token'
bigquery.credentials : ''   # leave empty to use VM IAM identity (ADC)
bigquery.project     : 'your-gcp-project-id'
bigquery.dataset     : 'checkmein'