Add Google Sheets as a data source

[db-tycoon-stephen]

Summary

Add Google Sheets as a first-class data source so users can pull a sheet/range into the local warehouse via tycoon data sources add → tycoon data sync.

Approach

Follow the established add-a-source process (source shim + CatalogEntry + dbt staging models). Per project convention, do not bundle the source — pull dlt's verified google_sheets source on demand via dlt init google_sheets duckdb.

• Wire google_sheets into the source registry + CatalogEntry (id, label, required config keys).
• Credentials: support service-account JSON (headless/cron-friendly) and document the OAuth path; surface a clear tycoon doctor row when creds are missing/invalid.
• Non-interactive flags so sources add can be driven without a TTY (consistent with Non-interactive flags for tycoon data sources add (unblocks online recipe doctests) #44).
• Generate a staging dbt model for the ingested sheet (header row → typed columns).
• Tests: mock the Sheets client — auth-ok, auth-fail, empty-sheet, range-subset.
• Docs: short recipe in the sources guide.

Notes

• Sheets has no native incremental key; first cut is full-refresh replace per sync.
• Range/tab selection should be a config option (default: first sheet, used range).

[db-tycoon-stephen]

Mirrored in Jira: PTC-22 (Epic: PTC-6 — Operator UX & demo recipes). Dev work continues here.