Turn new Firebase documents into Google Sheet rows

This is a deterministic logging pipeline. Use a Firebase poll trigger that fires whenever a new document is created in a specific Firestore collection (new_document). Firestore is the source of truth and a Google Sheet is the running, human-readable log, so non-technical teammates can see every new record without ever touching the database.

For each new document, capture the document id, its created timestamp (formatted as an ISO 8601 string), and a fixed, predefined set of fields. If the trigger payload does not already include the full document body, use Firebase Get Firestore Document to fetch it by its path before building the row.

Firestore wraps every field in a typed value envelope (stringValue, integerValue, doubleValue, booleanValue, timestampValue, mapValue, arrayValue, and so on). Flatten these into plain cell values: unwrap scalars to their raw value, render timestamps as ISO 8601, and serialize nested maps and arrays into compact, readable text so the spreadsheet stays clean. Keep a stable column order across every run so each column always means the same thing.

Append the resulting row to the Google Sheet using Google Sheets Append Values, writing one row per new document to a fixed range so rows accumulate as an append-only log. Never modify or delete existing rows, and never write back to Firestore.

Keep the mapping fully deterministic: the same document always produces the same row in the same column order. If a captured field is missing on a given document, write an empty cell for that column rather than shifting the other values, so the columns stay aligned.