MCP App Store
by KeenEthics
Foodics icon

Foodics MCP Server for ChatGPT

By Foodics

Overview

Connect your Foodics business to explore sales, orders, menu, inventory, customers and restaurant performance in English or Arabic. All tools are read-only. Connection requires a Foodics business owner or a user authorized to manage apps. Data access follows the connected business account.

How to connect

  1. ChatGPT: Open the Foodics listing in the ChatGPT app directory and choose Connect. ChatGPT listing Foodics exposes 74 tools here.

Tools

get_business_days_report

ChatGPT
Daily totals per business day (open/close times, net sales, order count) over a date range. Use for day-by-day operational summaries.

get_cost_adjustment_history_report

ChatGPT
Row-level log of manual cost adjustments over a date range. Use to audit who changed item costs and by how much.

get_cost_analysis_report

ChatGPT
Inventory cost breakdown over a date range. level is required — pass 'item' for per-ingredient costs or 'category' for category-level rollup. Use when you need COGS or cost-variance analysis.

get_customer

ChatGPT
Get a single customer by UUID. Rows carry a numeric gender code — read meta.enums for the labels.

get_customer_insights_report

ChatGPT
Lifetime summary for one specific customer (total spend, visit count, favourite items). Requires customer_id. No date range — covers lifetime history. Use get_sales_by_customer for date-bounded multi-customer analysis instead.

get_customer_orders

ChatGPT
Get order history for a specific customer. Order rows carry numeric status, type, source, delivery_status and discount_type codes — read meta.enums for the labels (note order_status decodes the order's own status, product_status decodes each line item's; they are different tables). Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

get_customer_points

ChatGPT
Get a customer's CURRENT loyalty points balance by UUID (the loyalty_balance on their record). For the earn/redeem history behind it, use get_loyalty_transactions instead.

get_drawer_operations_report

ChatGPT
Row-level cash drawer events (pay-in, pay-out, cash drop, open drawer, sales, returns) over a date range. Use to audit drawer activity or investigate discrepancies.

get_gift_cards_report

ChatGPT
Gift card activity over a date range (issued, redeemed, balances). Use for gift-card liability or redemption reporting.

get_inventory_control_report

ChatGPT
Opening stock, closing stock, consumption, and variance per item over a date range. Use for stock reconciliation or shrinkage analysis. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

get_inventory_history_report

ChatGPT
Row-level log of every inventory movement (adjustments, transfers, receipts) over a date range. Use when you need the full audit trail of stock changes. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

get_inventory_levels

ChatGPT
Get current live stock levels for ONE branch (raw per-item quantities on hand). Optionally filter to specific item UUIDs. For an aggregated point-in-time levels report across multiple branches with grouping, use get_inventory_levels_report() instead. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

get_inventory_levels_report

ChatGPT
Point-in-time stock quantities across branches — no date range required. Use for multi-branch inventory snapshots. For a single branch's live quantities use get_inventory_levels instead. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

get_loyalty_reward

ChatGPT
Get a single loyalty reward by UUID.

get_loyalty_rewards

ChatGPT
List available loyalty rewards. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

get_loyalty_transactions

ChatGPT
List loyalty transactions (the earn/redeem history). Optionally filter by customer or order UUID. For a customer's CURRENT balance, use get_customer_points instead. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

get_menu_engineering_report

ChatGPT
Menu engineering matrix over a date range — classifies each item as Star / Plow-horse / Puzzle / Dog based on popularity vs profitability. Use for menu optimisation decisions.

get_modifier_option

ChatGPT
Get ONE modifier option by UUID with every relation embedded, including allergens and nutrition_info — the only tool that returns those (search_modifier_options cannot include them). For other detail prefer search_modifier_options(ids=[...], include=[...]), which takes up to 50 at once. Here every relation is returned, so a missing one means none is set. Rows carry a numeric costing_method code — read meta.enums for the labels.

get_operations_report

ChatGPT
Live operational snapshot — active orders, free tables, online cashiers, last sync time per branch. No date range, real-time data only.

get_order

ChatGPT
Get a single order by UUID. Use include for full detail, e.g. include=["products.product","payments.payment_method","customer","branch"]. Order rows carry numeric status, type, source, delivery_status and discount_type codes — read meta.enums for the labels (note order_status decodes the order's own status, product_status decodes each line item's; they are different tables).

get_overview_report

ChatGPT
High-level KPI over a date range (net sales, order count, avg order price, top products/branches, etc.). Returns ONE KPI per call — the metric argument is required; call once per KPI you need. Use as the first stop for 'how are we doing?' questions. Whole-business only — this report accepts NO filters at all (no branches/branch_id); for per-branch figures use a sales_by_ report. There is also no gross_sales metric of any name here (see the `metric` param) — for gross revenue use a sales_by_ report's gross_sales column.

get_payments_report

ChatGPT
Payment totals by payment method over a date range. Use when you need to know how much was paid by cash, card, wallet, etc.

get_product

ChatGPT
Get ONE product by UUID with every relation embedded, including allergens and nutrition_info — the only tool that returns those (search_products cannot include them). For other detail prefer search_products(ids=[...], include=[...]), which takes up to 50 at once. Here every relation is returned, so a missing one means none is set. Rows carry numeric pricing_method, selling_method and costing_method codes — read meta.enums for the labels.

get_purchase_orders_report

ChatGPT
Row-level list of purchase orders over a date range. Use to audit procurement activity or find orders from a specific supplier.

get_purchasing_report

ChatGPT
Aggregated purchasing summary over a date range, grouped by branch or supplier. Use for spend analysis per supplier or per branch.

get_sales_by_branch_report

ChatGPT
Sales totals broken down by branch. Use when comparing branch performance (revenue, orders, discounts, tax). group_by slices each branch row further. IMPORTANT: the branch rows (no group_by, or a time, order-type or staff group_by) INCLUDE paid-modifier revenue and order charges in net_sales. Item group_bys split that money differently: group_by='product' or 'category' rows show the base item price only (no modifier revenue, no charges), group_by='modifier' rows carry the modifier revenue, group_by='combo' rows the combos. This report has no exclude_modifiers parameter, unlike sales_by_product/sales_by_category/sales_by_combo: passing one here is DISCARDED by the API before the query runs (verified live: 200 with unchanged numbers), so do NOT read that 200 as the flag having worked.

get_sales_by_cashier_report

ChatGPT
Sales totals broken down by cashier (the POS user who rang up the order). Use to evaluate cashier productivity or spot outliers.

get_sales_by_category_report

ChatGPT
Sales totals broken down by menu category. Use to see which menu sections drive the most revenue. net_sales INCLUDES paid-modifier revenue by default but never order charges: the default total plus sales_by_branch's charges equals sales_by_branch's total. exclude_modifiers=true gives the base item price only, which matches sales_by_branch with group_by='category'.

get_sales_by_charge_report

ChatGPT
Sales totals broken down by charge type (delivery fee, service charge, etc.). Use to analyse extra-fee revenue.

get_sales_by_combo_report

ChatGPT
Sales totals broken down by combo meal. Use to rank combo performance. net_sales INCLUDES paid-modifier revenue by default, which matches sales_by_branch with group_by='combo'; exclude_modifiers=true leaves the modifier revenue out.

get_sales_by_creator_report

ChatGPT
Sales totals broken down by order creator (staff member who opened the order, which may differ from the cashier). Use to track order-creation activity per employee.

get_sales_by_customer_report

ChatGPT
Sales totals broken down by customer — each row IS one customer (do NOT pass group_by='customer'). Use for top-spender ranking or loyalty-driven revenue breakdown. group_by slices each customer row further (e.g. by branch or product). Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

get_sales_by_delivery_zone_report

ChatGPT
Sales totals broken down by delivery zone. Use to compare delivery-area revenue and order volumes.

get_sales_by_discount_report

ChatGPT
Sales totals broken down by discount. Use to measure how much each discount drives revenue and how often it's applied.

get_sales_by_driver_report

ChatGPT
Sales totals broken down by delivery driver. Use to compare driver performance (orders delivered, revenue carried).

get_sales_by_modifier_option_report

ChatGPT
Sales totals broken down by modifier option (e.g. 'Extra cheese'). Use to see which add-ons are most popular and their revenue impact.

get_sales_by_order_source_report

ChatGPT
Sales totals broken down by order source. Rows carry a numeric order_source code — read meta.enums for the labels. Use to compare channel performance.

get_sales_by_order_type_report

ChatGPT
Sales totals broken down by order type. Rows carry a numeric order_type code — read meta.enums for the labels. Use to compare fulfilment-type revenue.

get_sales_by_product_report

ChatGPT
Sales totals broken down by product. Use for best/worst seller ranking, product revenue contribution. net_sales INCLUDES paid-modifier revenue by default but never order charges: the default total plus sales_by_branch's charges equals sales_by_branch's total. exclude_modifiers=true gives the base product price only, which matches sales_by_branch with group_by='product'.

get_sales_by_promotion_report

ChatGPT
Sales totals broken down by promotion. Use to measure promotion effectiveness and uplift.

get_sales_by_section_report

ChatGPT
Sales totals broken down by dining section (floor area / seating zone). Use to compare section utilisation and revenue.

get_sales_by_table_report

ChatGPT
Sales totals broken down by table. Use to compare table-level revenue and average covers.

get_sales_by_timed_event_report

ChatGPT
Sales totals broken down by timed event (scheduled menu/pricing windows such as happy hour or breakfast). Use to measure event-driven revenue.

get_shifts_report

ChatGPT
Row-level staff shift records (clock-in, clock-out, hours worked) over a date range. Use for payroll-input checks or labour-hour analysis.

get_speed_of_service_report

ChatGPT
Average timings per order stage (placed → kitchen received → kitchen done → closed) over a date range, all in seconds. group_by is REQUIRED — pick branch or creator.

get_taxes_report

ChatGPT
Tax collected by tax type over a date range. Use for tax reporting or VAT reconciliation.

get_tills_report

ChatGPT
Till/drawer session records over a date range — opening float, closing count, cash movements. Use for cash-management reconciliation. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

get_tips_report

ChatGPT
Tips collected over a date range, groupable by branch, order source, or order type. Use for tip-distribution or staff-incentive reporting.

get_transfer_orders_report

ChatGPT
Row-level list of inter-branch transfer orders over a date range. Use to audit transfer requests.

get_transfers_report

ChatGPT
Row-level list of completed inter-branch transfers over a date range. Filters use sender_branches and receiver_branches (not generic 'branches').

get_voids_returns_report

ChatGPT
Row-level log of voided items and returns over a date range — what was voided/returned, by whom, the reason, and whether it was marked as waste. Use for waste tracking or cashier-void audits. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_allergens

ChatGPT
Search allergens. Filterable by name and predefined flag. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_branches

ChatGPT
Search branches. name is free-text (substring match); reference is exact. Also filterable by online-order capability. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_categories

ChatGPT
Search menu categories. Filterable by name and reference. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_combos

ChatGPT
Search combo meals. Filterable by name, category, active status. Set full_detail=true for the whole tree in the same call: sizes, items, their options and products, with the combo prices. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_cost_adjustments

ChatGPT
Search cost adjustment transactions (corrections/overrides to inventory item costs). Filterable by branch, status, reference, business date range. Rows carry a numeric status code — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_coupons

ChatGPT
Search coupons. Filterable by name, code, discount, active status. Rows carry a numeric coupon_type code — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_customers

ChatGPT
Search customers. Filterable by name, email, phone, blacklist status, loyalty membership. Use it to find specific customers, e.g. 'customers named Ahmed' or 'the customer with this phone number', or to rank them with sort_by (e.g. the newest customer, or most orders ever). For a count, pass count_only=true (e.g. 'how many loyalty members do we have': is_loyalty_enabled=true, count_only=true). NOT for totals: top spenders or revenue per customer → get_sales_by_customer_report (e.g. 'top 20 customers this month'); one customer's lifetime spend and visits → get_customer_insights_report. Rows carry a numeric gender code — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — never present it as complete. Fetch the next batch (pass meta.next_starting_page as starting_page) only when you need the rows themselves; for totals, use the report named above instead of paging to the end. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_discounts

ChatGPT
Search discounts. Filterable by name, reference, percentage flag. Rows carry a numeric qualification code — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_inventory_categories

ChatGPT
Search INVENTORY item categories (raw-material groupings — distinct from MENU categories via search_categories). Filterable by name, reference. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_inventory_counts

ChatGPT
Search inventory count sheets (stock-takes). Filterable by branch, status, reference, business date range. Rows carry a numeric status code — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_inventory_items

ChatGPT
Search inventory items (raw materials / ingredients). Filterable by name, SKU, category. Use include=["ingredients","branches"] for recipe and stock detail. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_inventory_transactions

ChatGPT
Search inventory transactions (stock movements: purchasing, transfers, adjustments, production, consumption, waste, returns). Use it to look at specific movements, e.g. 'which transfers did branch X send yesterday'. NOT for totals: one report call replaces paging hundreds of rows. Consumption, variance, opening/closing stock per item → get_inventory_control_report (e.g. 'how much chicken did we use this month'); purchasing spend per supplier or branch → get_purchasing_report (e.g. 'how much did we buy from each supplier'); ingredient cost → get_cost_analysis_report (e.g. 'food cost by category': level='category'). count_only=true counts every type and status unless you filter them. Filterable by branch, transaction type, status, supplier, source purchase/transfer order, reference, business date range. Rows carry numeric type and status codes — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — never present it as complete. Fetch the next batch (pass meta.next_starting_page as starting_page) only when you need the rows themselves; for totals, use the report named above instead of paging to the end. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_modifier_options

ChatGPT
Search individual modifier options (the selectable choices inside a modifier group). Filterable by name, SKU, modifier group, tax group, active status, updated_after. Use include=["modifier"] to embed the parent group, "branches" for branch-level pricing, "ingredients" for recipe detail. Rows carry a numeric costing_method code — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_modifiers

ChatGPT
Search modifier groups. Filterable by name, reference, ready status. Use include=["options"] to see each group's choices and prices. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_orders

ChatGPT
Search orders (row-level — line items, payments, who handled it). Use it to look at specific orders, e.g. 'what was in order 1234' or 'this customer's orders yesterday'. NOT for totals: one report call replaces paging hundreds of orders. Sales per day or branch → get_sales_by_branch_report (e.g. 'sales per day this month': group_by='day_date'); payments by method → get_payments_report (e.g. 'how much was paid by card'); products or categories → get_sales_by_product_report / get_sales_by_category_report (e.g. 'top 10 products last week'); channels (API, call center, cashier) → get_sales_by_order_source_report (e.g. 'API vs cashier sales'); order types (dine in, pick up, delivery, drive thru) → get_sales_by_order_type_report; customers → get_sales_by_customer_report (e.g. 'top spenders'); discounts → get_sales_by_discount_report. count_only=true counts orders of EVERY status (pending, void, returned, draft…) unless you pass status (status=4 is Closed); for sales order counts use the reports above. REQUIRES a bound: a business date (exact or range) OR a specific identifier (reference / number / customer_id); branch/status/type alone are rejected. Lean by default — request includes only when you need line-item detail. Order rows carry numeric status, type, source, delivery_status and discount_type codes — read meta.enums for the labels (note order_status decodes the order's own status, product_status decodes each line item's; they are different tables). Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — never present it as complete. Fetch the next batch (pass meta.next_starting_page as starting_page) only when you need the rows themselves; for totals, use the report named above instead of paging to the end. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_products

ChatGPT
Search products. Filterable by name, SKU, barcode, category, tax group, active status, updated_after. Rows carry numeric pricing_method, selling_method and costing_method codes — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_promotions

ChatGPT
Search promotions. Filterable by name, type, active status. Rows carry numeric type, condition_type and reward_type codes — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_purchase_orders

ChatGPT
Search purchase orders (row-level orders with their line items — distinct from the aggregated get_purchase_orders_report). Filterable by branch, supplier, status, reference, business date range. Rows carry a numeric status code — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_suppliers

ChatGPT
Search suppliers. Filterable by name and code. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_tax_groups

ChatGPT
Search tax groups (bundles of taxes attached to products/branches). Filterable by name. Use include=["taxes"] to embed the individual rates. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_taxes

ChatGPT
Search individual taxes (named rates, e.g. VAT 15%). Filterable by name. Use include=["tax_groups"] to see which groups use this tax. Distinct from search_tax_groups, which lists bundles. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_transfer_orders

ChatGPT
Search transfer orders (row-level inter-branch/warehouse transfers with line items — distinct from the aggregated get_transfer_orders_report). Filterable by branch, warehouse, status, reference, business date range. Rows carry a numeric status code — read meta.enums for the labels. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

search_warehouses

ChatGPT
Search warehouses (central kitchens / item distributors). name is free-text (substring match); reference is exact. Rows are lean: a related object (category, branches, options, …) is present only when named in include — a missing key means not requested, not empty. To look up known records by UUID, pass ids=[...] (up to 50, live or deleted) with the include you need, in one call. Returns up to 200 items per call. If meta.has_more is true, THE RESULT IS INCOMPLETE — do not summarize, count, chart, or answer from it yet. Call this tool again passing meta.next_starting_page as starting_page to fetch the next batch, and repeat until meta.has_more is false before presenting any answer. If meta.rate_limited_retry_after is set, the batch stopped at the rate limit: wait that long before fetching the next one.

Capabilities

Analyze Restaurant SalesExplore Orders And PaymentsRead Menu And Inventory DataReview Branch Performance

Example Prompts

Click any prompt to copy it.

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/foodics"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/foodics)