Migrating from SellerBoard

You are using DataDoe MCP. Produce a SellerBoard-style P&L summary.
Include parent rows only and one Total for the date range.
Match Reports → Profit & Loss for the Seller and dates below.
Do not include daily, weekly, monthly, or comparison columns.

Inputs:
- seller name: {{seller_name}}
- from: {{from}}   (YYYY-MM-DD, inclusive)
- to: {{to}}       (YYYY-MM-DD, inclusive)

Do not use SQL, BigQuery, or the DataDoe web app.
Use only the tools named below.
Follow this order.

1) Tool: sellers_and_vendors_list
   Arguments (exact keys):
   {
     "query": "{{seller_name}}",
     "page": 1,
     "pageSize": 10
   }
   Pick the Seller Central row whose display name matches {{seller_name}}.
   Save its id as sellerOrVendorId (UUID).

2) Tool: exports_sources_get
   Arguments (exact keys):
   {
     "sellerOrVendorIds": ["<sellerOrVendorId from step 1>"],
     "query": "Profit by Date",
     "page": 1,
     "pageSize": 8
   }
   Select the source where table is amazon_profit_by_date and the user-facing name is "Profit by Date".
   Save sourceId from that object.
   If enabled is false, stop and tell the user to enable the table.

3) Tool: exports_source_get
   Arguments (exact keys):
   {
     "sellerOrVendorIds": ["<sellerOrVendorId from step 1>"],
     "sourceId": "<sourceId from step 2>",
     "page": 1,
     "pageSize": 40
   }
   This returns at most 40 columns. If columnsMeta.hasNextPage is true, the column list is incomplete. Call the same tool again with page 2, then further pages, until hasNextPage is false. Confirm columns such as total_sales, total_units_sold, ad_spend, profit, and total_sessions exist across those pages.

4) Tool: exports_create
   Create one total for {{from}} to {{to}} on amazon_profit_by_date.
   Use the exact argument keys below.
   Values for aggregations.column must be source columns.
   Values for columns must be the aliases:
   {
     "sellerOrVendorIds": ["<sellerOrVendorId from step 1>"],
     "sourceId": "<sourceId from step 2>",
     "from": "{{from}}",
     "to": "{{to}}",
     "outputType": "JSON",
     "limit": 1,
     "columns": [
       "sales",
       "sales_tax",
       "sales_tax_missing_items",
       "units",
       "refunds",
       "advertising_cost",
       "refund_cost",
       "amazon_fees",
       "cost_of_goods",
       "net_profit",
       "sessions"
     ],
     "aggregations": [
       { "column": "total_sales", "aggregation": "sum", "alias": "sales" },
       { "column": "sales_tax", "aggregation": "sum", "alias": "sales_tax" },
       { "column": "sales_tax_missing_items", "aggregation": "sum", "alias": "sales_tax_missing_items" },
       { "column": "total_units_sold", "aggregation": "sum", "alias": "units" },
       { "column": "refund_count", "aggregation": "sum", "alias": "refunds" },
       { "column": "ad_spend", "aggregation": "sum", "alias": "advertising_cost" },
       { "column": "refund_cost", "aggregation": "sum", "alias": "refund_cost" },
       { "column": "total_fees", "aggregation": "sum", "alias": "amazon_fees" },
       { "column": "cogs_total", "aggregation": "sum", "alias": "cost_of_goods" },
       { "column": "profit", "aggregation": "sum", "alias": "net_profit" },
       { "column": "total_sessions", "aggregation": "sum", "alias": "sessions" }
     ]
   }
   Do not set groupBy, dateInterval, filters, skip, or orderByColumn.

   Report row → export field:
   Sales → sales (UI: shipped item price plus buyer-paid shipping, minus shipping promotions; unshipped orders are excluded)
   Sales tax → sales_tax (show as a negative cost; profit already subtracts this positive cost once)
   Items missing tax data → sales_tax_missing_items
   Units → units
   Refunds → refunds
   Promo → always 0 (not in this table)
   Advertising cost → advertising_cost (reported ad spend; no blanket VAT uplift is applied)
   Shipping costs → n/a (buyer-paid shipping, minus shipping promotions, stays inside sales)
   Giftwrap → n/a (not on Profit by Date)
   Refund cost → refund_cost (settlement refund money, already signed; the report also includes the product cost of returns Amazon received)
   Amazon fees → amazon_fees
   Cost of goods → cost_of_goods
   Gross profit / Net profit → net_profit
   Estimated payout → n/a
   Sessions → sessions

5) If the status from step 4 is not already COMPLETED, tool: exports_get
   Arguments:
   { "exportId": "<id from step 4>" }
   Poll every 5 seconds until status is COMPLETED. Stop on ERROR or BLOCKED_NO_TOKENS.

6) Tool: exports_raw_download
   Arguments:
   { "exportId": "<same id>" }

Present the parent rows in the following report order.
Do not add other rows.
Show costs as negative values.
Show sales, units, and sessions as positive values.

- Sales = sales
- Units = units
- Refunds = refunds
- Promo = 0
- Advertising cost = if advertising_cost > 0 then -advertising_cost else advertising_cost
- Shipping costs = n/a
- Giftwrap = n/a
- Refund cost = refund_cost (already signed)
- Amazon fees = if amazon_fees > 0 then -amazon_fees else amazon_fees
- Cost of goods = if cost_of_goods > 0 then -cost_of_goods else cost_of_goods
- Sales tax = if sales_tax > 0 then -sales_tax else sales_tax
- Items missing tax data = sales_tax_missing_items
- Gross profit = net_profit
- Net profit = net_profit
- Estimated payout = n/a
- TACOS = abs(Advertising cost) / Sales   (if Sales is 0, null)
- % Refunds = Refunds / Units   (if Units is 0, null)
- Margin = Net profit / Sales   (if Sales is 0, null)
- ROI = Net profit / abs(Cost of goods)   (if Cost of goods is 0, null)
- Active subscriptions = 0
- Sessions = sessions
- Unit session percentage = Units / Sessions   (if Sessions is 0, null)

State that this is the Profit by Date total for {{from}} to {{to}}.
Explain that sales tax is shown as a separate cost where Amazon prices include it, that the profit field already subtracts it, and that cost of goods sold (COGS) uses unit costs uploaded to DataDoe.