Daily portfolio snapshot logger in Google Sheets

Build me a deterministic daily portfolio snapshot logger. This is a pure data pipeline with no AI step, no narrative output, just structured data transformation and logging so a Google Sheet builds a clean daily history I can chart from.

Trigger: cron, every US trading weekday (Monday through Friday) at 4:15pm America/New_York, after the market close.

Step 1. Use google-sheets Get Values to read my portfolio sheet. The range covers a header row plus a body where each row has two columns: ticker symbol and share count. Stock tickers are plain (AAPL, MSFT). Mutual fund tickers are prefixed with M: (for example M:VFIAX). The spreadsheet ID and the source range should be workflow inputs so I can point it at any sheet.

Step 2. Partition the rows into two groups: company tickers (no prefix) and mutual fund tickers (M: prefix). Drop blank rows.

Step 3. For company tickers, call ycharts Get Company Data Points in batches of up to 100 symbols per request, asking for the latest closing price and the 1-day percent change. For mutual fund tickers, call ycharts Get Mutual Fund Data Points the same way, also in batches of up to 100. Use the correct YCharts metric codes for closing price and 1-day percent change. Per YCharts common docs: max 100 symbols per request, mutual funds keep their M: prefix in the call.

Step 4. For every holding, compute position value (shares times price) and dollar change for the day (shares times price times percent_change/100, or equivalently value minus prior value implied by the percent move). Round to two decimals.

Step 5. Use google-sheets Append Values to append one row per position to the Daily Snapshot tab in the same spreadsheet. Row schema: date (YYYY-MM-DD in ET), ticker (with M: prefix preserved for funds), shares, close_price, percent_change_1d, position_value, dollar_change_1d. Use USER_ENTERED so numbers and dates are stored as native types, not strings. Use insertDataOption INSERT_ROWS so the appended rows do not overwrite existing data.

Error handling: if YCharts returns an error in meta.status or a 414 due to too many symbols, log the failing batch and continue with the rest. Do not block the entire run on one bad ticker. If a single ticker comes back without a price (delisted, halted, no close yet), skip it but write a row with an explanatory note in a status column so I can see it in the log.

No AI step. No Slack or email output. The Google Sheet IS the output. Keep the code workflow deterministic and idempotent enough that re-running the same day appends a duplicate batch rather than corrupting prior history.