get_business_profile
ChatGPTEverything Nuira knows about this company: its stated economics and targets (blended margin, target ROAS, target CAC, monthly revenue goal, restock lead time) with where each came from; the things it has been told about how the business works (decision rules, seasonality, operations, strategy), including recommendation_feedback — what the operator said about a specific recommendation, such as a product being a one-off promo, which explains why Nuira stopped suggesting it; and the suppliers, sales channels, locations and product lines it depends on. Use for any question about the business itself, what Nuira knows, or where a figure came from. Its cogsCoverageAllTime and marginBasisAllTime describe the whole order history, NOT the period on screen: for how much of the current period's revenue has a known cost, use get_summary or get_product_performance and quote that figure instead.
get_channels
ChatGPTSpend, revenue and ROAS broken down per marketing channel (Meta, Google Ads, Klaviyo). Use when the question is about which channel is responsible for something. Optional from and to (inclusive YYYY-MM-DD, both or neither, never after today) read another span instead of the default period — a named month, a week, two dates the operator gave.
get_contribution_breakdown
ChatGPTThe full descent from net revenue to contribution for the current period: net revenue, cost of goods sold, gross profit, ad spend and contribution after media, each with its share of revenue. This is what the Contribution Summary card on the Finance page shows. Use for any question about margin structure, COGS, gross profit, or what is left after costs — get_summary does NOT carry COGS or gross profit. Optional from and to (inclusive YYYY-MM-DD, both or neither, never after today) read another span instead of the default period — a named month, a week, two dates the operator gave.
get_data_status
ChatGPTWhich data sources are connected for this brand and whether their data is complete: per source its status, freshness, last successful sync and, while a first import is running, how far it has got. Call this before reporting figures to someone new to Nuira, and whenever other tools return zeros or empty lists — a store with nothing connected or mid-import has real sales Nuira has not read yet, so its zeros are missing data, never the brand's figures.
get_email_performance
ChatGPTKlaviyo flows and email campaigns for the current period: revenue, sends, open rate and click rate per flow and campaign. Use when asked about email or SMS performance, or which flow earns the most. Optional from and to (inclusive YYYY-MM-DD, both or neither, never after today) read another span instead of the default period — a named month, a week, two dates the operator gave.
get_fact_history
ChatGPTHow the figures Nuira holds about this business have changed over time: each value, when it took effect, when it was superseded, where it came from, and the document and snippet it was read from. Use when asked what a figure used to be, when it changed, or why Nuira believes it.
get_inventory_health
ChatGPTStock levels per SKU, banded by how long the stock lasts at the current sales rate: out of stock, critical (under a week), at risk (one to two weeks) and healthy. Includes the SKUs closest to running out, with days of cover and units left. Use for any question about stock, restocking, or what will run out. The untracked count is SKUs whose stock Shopify does not track — they are in no band and are not a problem, they are simply uncounted.
get_knowledge_coverage
ChatGPTHow much Nuira actually knows about this business, per area, as a share from 0 to 1: margins is the share of revenue whose orders carry a known supplier cost (this is the answer to any question about COGS coverage — how much of the business has costs recorded), customers the share of revenue tied to an identified customer, catalogue and channels whether products and ad sources are connected, documents how many have been ingested. Use for "what do you know about us", "how good is your data", and for how complete the COGS or cost data is. A null coverage means that area has nothing to measure yet, which is different from a coverage of 0.
get_margin_trend
ChatGPTGross margin and margin-after-spend month by month, so a figure can be placed against its own history. Use when asked whether a margin is normal, when it changed, or how it compares with earlier periods.
get_money_flow
ChatGPTWhich marketing channels are credited with revenue on which products, as a network of channel and product nodes joined by attributed revenue and order count. Use when asked what a channel actually sells, or which products a channel's spend is earning.
get_product_categories
ChatGPTRevenue, units and margin grouped by product category for the current period. Use when asked which category or product type drives the business. IMPORTANT: when suppressed is true, too few products carry a category for any split to be trustworthy — say the store's categories are mostly unset and give no breakdown, rather than reporting the rows. Optional from and to (inclusive YYYY-MM-DD, both or neither, never after today) read another span instead of the default period — a named month, a week, two dates the operator gave.
get_product_performance
ChatGPTHeadline catalogue figures for the current period: product revenue, units sold, average order value, gross margin and how many SKUs are out of stock, each with the change against the previous period. Use first for any question about how the catalogue as a whole is doing. Gross margin comes back null when supplier costs are missing for some of the sales — say the margin is unknown rather than reporting it as zero. Optional from and to (inclusive YYYY-MM-DD, both or neither, never after today) read another span instead of the default period — a named month, a week, two dates the operator gave.
get_revenue_by_hour
ChatGPTRevenue per hour of the local day, over the last 7 days. Use when the question is about time of day — which hours sell, when a spike happened within a day, when to schedule something. REVENUE ONLY: ad spend arrives as a date with no hour attached, so there is no hourly spend, margin or ROAS to report. If asked for those by hour, say the data does not exist at that resolution rather than dividing a daily figure across the day. Hours are the project's own local hours, and a day on which the clocks change has 23 or 24 buckets rather than 24.
get_sales
ChatGPTSales performance over the last 30 days broken down by campaign and by channel, with how many orders had more than one touchpoint and how last-touch attribution compares with other models. Use for questions about campaigns, attribution, or which channel really earned a sale. For what the catalogue itself sold, use get_top_products or get_product_performance instead.
get_sales_breakdown
ChatGPTShopify's own Total sales breakdown for the current period: gross sales, discounts, returns, net sales, shipping and taxes — the same lines, with the same definitions, as the merchant's Shopify admin. Use this whenever the question is about revenue in the merchant's own terms, or when they say our number disagrees with Shopify. IMPORTANT: netSales is Shopify's revenue figure and EXCLUDES tax and shipping; orderTotal is the order-total basis the rest of Nuira still reports and INCLUDES both, which is why the two differ — say so rather than picking one silently. Returns count in the period the refund was issued, not the period of the order it reverses. When complete is false, some orders in the range predate these figures and every line understates: say that instead of quoting the totals as final. Optional from and to (inclusive YYYY-MM-DD, both or neither, never after today) read another span instead of the default period — a named month, a week, two dates the operator gave.
get_series
ChatGPTDaily revenue and spend over the current period. Use when the question is about when something changed, or about a trend or spike. Optional from and to (inclusive YYYY-MM-DD, both or neither, never after today) read another span instead of the default period — a named month, a week, two dates the operator gave.
get_summary
ChatGPTHeadline figures for the current period: revenue, ad spend, contribution and blended ROAS, each with the change against the previous period, plus new and returning customers, order count, sessions and conversion rate. Use this first for any question about overall performance, and for how many customers are new or returning. Optional from and to (inclusive YYYY-MM-DD, both or neither, never after today) read another span instead of the default period — a named month, a week, two dates the operator gave.
get_top_products
ChatGPTThe ten products that earned the most in the current period, with revenue, units, cost (what those units cost), grossProfit (revenue minus that cost), margin and the change against the previous period. Each row also carries revenueShare — its fraction of these ten products' revenue — and rankedRevenue is their combined total, so a breakdown by product needs no arithmetic of your own. Use when asked which products sell best, which are growing or slowing, which are most profitable, or for a breakdown of revenue by product. A null cost, grossProfit or margin means costs are missing for that product, not that it earns nothing — never treat a null as zero and never fill one in by subtracting. If truncated is true some orders had more line items than Shopify returns at once, so the split is partial — say so. Optional from and to (inclusive YYYY-MM-DD, both or neither, never after today) read another span instead of the default period — a named month, a week, two dates the operator gave.
get_vertical_benchmarks
ChatGPTTypical margin and ROAS ranges for the commerce vertical this store's catalogue looks like. These are INDUSTRY REFERENCE values for comparison — they are NOT this brand's own figures and must never be reported as such. Use only to say whether one of the brand's own numbers is normal for its category. Returns available=false when the catalogue cannot be classified.
list_insight_episodes
ChatGPTHow a detected change unfolded over time — the run of days an anomaly persisted, rather than the single snapshot list_insights returns. Use when asked how long something has been happening, or whether a change is still going on.
list_insights
ChatGPTDetected week-over-week changes and anomalies that are still going on, most impactful first, each with the two weeks it compares. Use when asked what changed, or why something moved — and check each change's window against the period you are describing before connecting them.
list_recommendations
ChatGPTThe ranked recommendations currently open for this project, with expected value and confidence. Use when asked what to do, or what the biggest opportunity is.
get_business_profile
Claudeget_channels
Claudeget_connect_links
Claudeget_contribution_breakdown
Claudeget_data_status
Claudeget_email_performance
Claudeget_fact_history
Claudeget_inventory_health
Claudeget_knowledge_coverage
Claudeget_margin_trend
Claudeget_money_flow
Claudeget_product_categories
Claudeget_product_performance
Claudeget_revenue_by_hour
Claudeget_sales
Claudeget_sales_breakdown
Claudeget_series
Claudeget_summary
Claudeget_top_products
Claudeget_vertical_benchmarks
Claudelist_insight_episodes
Claudelist_insights
Claudelist_recommendations
Claude
