add_ideas_to_pipeline
ChatGPTPuts Product Ideation results (ASINs from search_product_ideas) into the Product Pipeline so they can be validated: mode separate = one project per ASIN; together = one project with all of them as competitors (give it a name). Existing projects are returned, never duplicated. Then judge each project with get_project_criteria.
add_negative_keywords
ChatGPTAdd a search term as a negative keyword to one or more campaigns in Sotrics' negative-keyword plan (exact, phrase or ASIN). This is recorded in Sotrics and the action log; it is not pushed to Amazon Ads (Ads writes are not enabled). Campaign names from get_ppc_campaigns.
add_sourcing_product
ChatGPTCreates a Supplier Tracker product for a product being sourced, linked to its pipeline project when there is one (the price and Amazon fees then come from the project; typed values win). Returns the sourcing product id for add_supplier_quote and start_sourcing_automation.
add_supplier_quote
ChatGPTAdds one supplier (contacted or quoted) to a Supplier Tracker product: supplier details, MOQ, EXW / FOB / DDP prices, add-ons, carton data, freight, tariff, lead time and sample notes. Sotrics computes landed cost, margin, ROI and the score itself (get_supplier_quotes). Record only what the supplier actually said.
add_tracked_keywords
ChatGPTStart tracking organic and sponsored rank for keywords of one product (ASIN). Paused keywords added again are resumed. Uses the account's keyword allowance.
decide_research_run
ChatGPTThe one decision of a Research Automation run that reached its decision stage (get_research_runs): approve = continue to trademark / patent and the sourcing hand-off; reject; or hold. Give the reason as note.
get_action_log
ChatGPTRecorded actions on the account (price/listing changes detected from Amazon, PPC changes, decisions, comments), newest first.
get_agent_learnings
ChatGPTLessons, notes, tests and brand profiles the Sotrics AI agent team saved on this seller account since a date (default the last 7 days), oldest first, each with the agent, scope (account / brand / listing), brand, ASIN and evidence. Used by the company brain's nightly review to file what the agents learned into the brain. Read only; these are agents' conclusions, not decisions.
get_alerts
ChatGPTOpen alerts for the account (stockouts, rank drops, listing changes, Buy Box loss ...).
get_brain_context
ChatGPTThe company brain of this seller account (its private GitHub repository of written rules, decisions, owners, suppliers, policies and where numbers live): the one-page context with links to every home note, plus the folder map. Read this FIRST for any question about the company itself; then open the note with get_brain_note and cite its path. Live sales, ads, stock, rank and profit are never in the brain. Team-area notes only.
get_brain_note
ChatGPTOne note of the company brain by its path (for example 08-Supply-Chain/suppliers.md or 02-Policies/POL-001-gates.md): header (type, status, owner, confidence, freshness) and body with labelled Fact / Inference / Proposal lines and their sources. Paths come from get_brain_context or search_brain. Owner-only notes are never available here.
get_cash_flow
ChatGPTCash-flow plan: expected Amazon payouts, supplier payments, freight and other planned items by month.
get_competitor_changes
ChatGPTDetected competitor listing changes (price, title, images, bullets, coupons ...) with dates.
get_competitor_history
ChatGPTDated readings of one competitor of a research project: price, best seller rank, rating, review count, "bought in past month" sales and revenue. Sources: Amazon sales history (up to 2 years: daily for 90 days, weekly before) and the Sotrics extension; each reading names its source. Sales are minimums.
get_competitors
ChatGPTTracked competitor ASINs with price, BSR, rating, reviews and changes over the range.
get_differentiation_brief
ChatGPTThe Differentiation brief of a research project: market level (1 me-too to 4), the ladder gaps, the Level 2 offer variants with their gates (25% margin / 150% ROI on landed cost), voice-of-customer openings and table stakes, the chosen offer and the hand-off spec for sourcing (pack size, price, target unit cost, what to ask suppliers for). Read this before sourcing a product.
get_expenses
ChatGPTAccount-level Amazon expenses and fees (storage, subscriptions, reimbursements, other Finances lines).
get_external_traffic
ChatGPTTraffic that reaches an Amazon listing or brand from outside Amazon, as last looked up in Sotrics: estimated external visits per month (low / mid / high), share of total traffic, each referring page (deal sites, coupon and cashback, social, YouTube, influencer storefronts, blogs and reviews, other retailers) with type, estimated visits, confidence and first / last seen, a 12-month timeline with BSR and spike markers, a cross-check against search traffic and BSR sales, and Google search demand for the brand and main keyword. Every figure is an estimate. Read only: it never buys new data.
get_inventory
ChatGPTFBA/FBM inventory per SKU: available, inbound, reserved, days of supply, sales velocity, stockout risk and reorder suggestions.
get_ip_checks
ChatGPTThe saved USPTO trademark and patent checks of a research project (verdicts, similar marks, classes, patents with kind, status and estimated expiry) and the suggested terms to check. Run a new check with run_trademark_check or run_patent_check.
get_keyword_analysis
ChatGPTMaster keyword list of a research project (reverse-ASIN keyword data or imported keyword research files): every keyword with search volume, relevancy, competing products, word count, average rank and each competitor's organic/sponsored rank; competitor market capture and strength; launch keywords; root words; segments; data sources and dates. Search volumes are provider estimates.
get_keyword_harvest
ChatGPTSearch terms that earn their own exact campaign (500+ searches with orders, or 3+ orders in discovery campaigns) with the suggested budget and bid, per the account's harvest thresholds. Sample search-term data until the Ads API is connected.
get_keyword_history
ChatGPTHistory of one keyword of a research project: each competitor's organic and sponsored rank over time (daily Rank tracker checks + each keyword data load), search volume estimates over time, and Amazon's weekly searches (Search Query Performance) when the account's own products appeared for it.
get_launch_plan
ChatGPTThe launch preparation of a research project: launch keywords with relevance and objective, normalized keyword groups, batches by root and the campaign plan Sotrics builds from them (one campaign per keyword for the big ones, grouped for small ones; names, budgets, start dates, overlaps). Campaigns are created from the Campaign builder page by a person.
get_listing_readiness
ChatGPTReads ONE listing live from Amazon (Listings Items API) and scores its readiness: title, bullets, description, backend terms, required and optional attribute fill, image count, review topics covered, Amazon's own errors; checks that cannot be made are marked unknown. Needs the Product Listing role on the seller's SP-API app. Drafting and sending changes happen on the Listing assistant page.
get_listing_skus
ChatGPTThe account's own listings as stored from Amazon: SKU, ASIN, title and status. Pick a SKU for get_listing_readiness.
get_negative_keywords
ChatGPTSaved negative keywords per campaign and the Never keyword list with its deployment status.
get_ppc_benchmarks
ChatGPTCampaign performance against the account's PPC benchmark register (target and ceiling ACOS per objective, SV tags, wasted spend, negative-keyword candidates, diagnosis).
get_ppc_campaigns
ChatGPTCampaign-level PPC management view: campaigns with budget, spend, sales, ACOS, objective and targets. Needs the Amazon Ads API.
get_ppc_overview
ChatGPTAdvertising overview: spend, ad sales, ACOS, TACOS, ROAS, CPC, CTR, CVR by campaign type and product. Needs the Amazon Ads API; the response says when it is not connected yet.
get_pricing
ChatGPTCurrent prices, Buy Box status, price type (regular/deal/custom), unit economics and price ladder per product.
get_product_costs
ChatGPTThe saved landed cost per SKU: product cost, freight, shipping / prep, other, total landed cost and when it was updated. Every profit figure reads these; a SKU without a cost shows no profit. Change them with update_product_costs.
get_product_health_checkup
ChatGPTSKU optimization / product health checkup (Combined Rank analysis): net revenue, CM1/CM2/CM3, fee % on net revenue, ROC (recovery of capital = net revenue ÷ (net revenue − CM)), 1000-unit turnover, quintile ranks 1–5 per metric, Combined Rank (average of the selected metrics), priority band (protect & push / near-winner / low performer / bleeder) and keep/fix/drop call per product. Read it at parent level first; a rank is a priority inside the catalog, not a quality grade.
get_product_pipeline
ChatGPTProduct research projects (from the Sotrics Chrome extension) with status, market size, margins, competitors and launch budget.
get_products
ChatGPTEvery product (ASIN/SKU) with sales, units, orders, refunds, fees, cost, profit, margin, ROI, sessions, conversion, stock and lifecycle stage for a date range.
get_profit_and_loss
ChatGPTProfit and loss statement: one column per day/week/month/quarter plus a total (sales, refunds, Amazon fees, cost of goods, advertising when connected, gross and net profit, margin, ROI) and per-product P&L for the range.
get_project_criteria
ChatGPTJudges one research project against Sotrics' product criteria, exactly like the Market Analysis checker: overall pass / fail / incomplete, each tier, the competitors that QUALIFY (ASIN, price, monthly sales, reviews) and the manual checks left for a person. preset: basic (Sotrics Basic: 1+ seller at $10,000+ a month, under 350 reviews, FBA fee <= 25% and Amazon fees < 40% of the price, price above $25) | standard (Sotrics Standard: 2+ sellers at $10,000+ a month, under 1,000 reviews, FBA <= 25%, price above $18) | low_ticket (Low ticket products: 4+ competitors with 100-1,000 reviews, 300-1,000 units a month and a price under $15, plus 4+ under 200 reviews) | default, a saved criteria set id, or empty for the project's active criteria. tolerance: % the limits may be missed by (0 = strict; the owner allows 20 when a niche misses narrowly, never more than 50): judge strict first, then at 20, and say which one passed. A product is validated only when this says pass.
get_purchase_orders
ChatGPTPurchase orders with supplier, items, quantities, costs, payment and delivery status.
get_rank_tracker
ChatGPTTracked keywords per ASIN with daily organic and sponsored rank, search volume and changes.
get_research_project
ChatGPTOne product research project: settings, costs, fee calculator, cash-flow plan, status history and every competitor (price, sales, revenue, BSR, reviews, fees). Keyword data: get_keyword_analysis; competitor history: get_competitor_history.
get_research_run
ChatGPTOne Research Automation run: stages, every product with its checks (revenue, reviews, price, size, fees), segments, offers, landed cost, flags, trademark / patent results and the event trail.
get_research_run_brief
ChatGPTUse this when the person asks WHY a researched product was chosen, what the strongest reasons are, what the risks are, or asks you to check / re-check its numbers. For one Research Automation run it returns the ranked reasons with their numbers and sources (POE conversion and return rate, market concentration, criteria, the chosen market segment and its keyword result, what sells best, the offer with profit / margin / ROI and the competitor price points), the risks, what is still unknown, and checks: the derived numbers (landed cost, gross profit, margin, ROI, unit price, segment share, brand share, products passing) recomputed from their raw inputs and compared with the stored values (ok true / false / null = cannot be re-checked). Quote these figures exactly; never do the arithmetic yourself and never invent a number. Say plainly which checks matched, which differ and which could not be checked. Needs the run id from get_research_runs.
get_research_runs
ChatGPTResearch Automation runs of the account (keyword -> validated products with offer and landed cost): each run's keyword, marketplace, stage, status, qualifying products and whether it waits for a decision. status: open (default) | all | a stage status.
get_research_setup
ChatGPTResearch Automation set-up of the account: which providers are connected (search page capture, Keepa, AI), the thresholds (basic check, segments, keywords, offer gate), today's usage against the daily run limit and the run counts by status. Read this before starting runs.
get_restock_forecast
ChatGPTSales forecast for one ASIN (velocity, confidence, 80% range, stockout date, reorder date and quantity) using the account's forecast settings.
get_returns
ChatGPTReturns and refunds per product with return reasons and themes.
get_review_analysis
ChatGPTReview analysis of a research project from the reviews imported for its competitors: star split, themes (critical / neutral / positive), repeated phrases, customer requests and rule-based product ideas. Reviews arrive only by import on the Review analysis tab; without an import the answer says so.
get_review_intelligence
ChatGPTWhat buyers say about one product (Amazon Customer Feedback API, stored by Sotrics): top positive and negative review topics with share of reviews, star-rating impact, subtopics and real snippets; the category (browse node) benchmark for all products and the top 25%; category return reasons; 6-month topic trends; customer profile (who, when, where, what for, why bought, wishes) from reviews; ranked gap findings; optional competitor comparison; and the latest AI voice-of-customer report. Read only: it never calls Amazon.
get_sales_by_region
ChatGPTOrders and sales by ship-to state/region and country.
get_sales_forecast_plan
ChatGPTThe Sales Forecast plan of one product (child ASIN, or p:<parent id> for a parent): per month the seasonal coefficient, average daily units, expected units and revenue, stock from today with open POs, suggested orders, re-order and ship dates, profit (CM1-CM3, TACoS) and 12 months of cash. Without product: the list of products to pick from.
get_sales_overview
ChatGPTHeadline business numbers for comparable periods (sales, orders, units, refunds, Amazon fees, profit, margin, sessions, conversion) with the previous period for comparison. Choose the period grain.
get_search_query_performance
ChatGPTAmazon Brand Analytics Search Query Performance: queries with search volume, impression/click/cart/purchase share for the brand, per week.
get_search_terms
ChatGPTCustomer search terms from PPC with clicks, spend, orders, sales, ACOS and CVR, including wasted spend. Needs the Amazon Ads API.
get_sourcing_automation
ChatGPTThe Sourcing automation (Alibaba outreach through the owner's own account and worker): every product with its automation status, wave, suppliers found and the gate (3 suppliers at 25% margin / 150% ROI), open approvals, sender accounts and whether the worker is connected. Per product: get_sourcing_product.
get_sourcing_product
ChatGPTOne product's sourcing automation: every supplier against the gate, the waves, the conversations (phase, last message, who writes next, human takeover) and the approvals waiting for the owner.
get_strategy_decoder
ChatGPTHow a competitor family gets its sales, from data already stored (Keepa readings, keyword ranks, rank checks): traffic per variation (organic vs sponsored), estimated units, unit economics and ad spend, the loss-leader check, ad aggression, keyword map and the 12-month promotion playbook. Loading new data for an ASIN is done on the page by a person.
get_supplier_quotes
ChatGPTOne Supplier Tracker product with every supplier quote (status, MOQ, EXW / FOB / DDP prices, add-ons, cartons, freight, tariff, lead time, samples, communication score) and Sotrics' maths per quote: landed cost per unit, margin %, ROI %, score /100, failed checks; plus the summary (target unit cost 1/5-1/7 of the price, best landed, best ROI, the suggested shortlist of 3). Hard floors: 25% margin and 150% ROI.
get_supplier_tracker
ChatGPTEvery product being sourced in the Supplier Tracker: name, linked pipeline project, selling price, Amazon fees, how many suppliers were contacted and quoted, the selected supplier. Details and the quotes of one product: get_supplier_quotes.
get_traffic_score
ChatGPTEstimated monthly search impressions per competitor of a research project (keyword search volume x visibility of the rank; organic and sponsored apart), per parent, niche totals, head / mid / long-tail split, potential keywords and the 7-day movement. Uses stored keyword data only.
get_trends
ChatGPTPer-product trend table of key metrics (sales, units, profit, margin, sessions, conversion ...) across periods, to spot products that are growing or declining.
get_work_orders
ChatGPTThe account's work orders: goal, department, status (planned = waiting for the owner, running, done, failed, declined, expired), steps with their results, tasks for a person and the outcome checkpoints. Big actions the team proposes land here.
list_accounts
ChatGPTSeller accounts this Sotrics login can open, with marketplaces, Amazon connection status and which is the default. Call this first.
pause_sourcing_automation
ChatGPTPauses a running Sourcing automation: no new suppliers are contacted until it is resumed. Conversations and progress are kept. It never agrees payment, samples or terms (those wait for the owner as approvals), and the 25% margin / 150% ROI gate can be raised, never lowered.
record_action
ChatGPTWrite one entry in the account's action log: a change that was made (for example a price change, a listing update, a bid change) with before/after values and the reason. Entries can never be edited or deleted afterwards. Categories: PPC, LISTING, INVENTORY, OTHER. actionType must belong to the category; valid types: campaign_created, campaign_archived, campaign_enabled, campaign_paused, ad_group_created, ad_group_paused, keyword_added, keyword_removed, bid_increase, bid_decrease, budget_increase, budget_decrease, negative_added, negative_removed, placement_adjustment, targeting_change, bidding_strategy_change, automation_rule_run, automation_safety_change, automation_config_change, benchmark_change, campaign_objective_change, product_ppc_profile_change, never_list_change, launch_keywords_change, campaign_plan_prepared, ppc_ai_decision, title_change, bullets_change, listing_attribute_change, description_change, main_image_change, secondary_image_change, aplus_change, price_change, sale_price_change, coupon_change, coupon, deal_started, deal_ended, lightning_deal, variation_change, backend_keywords_change, listing_status_change, inventory_change, restock, inbound_shipment, replenishment_change, inventory_alert, stockout, po_status_change, po_received, note, health_checkup_decision, meeting_action, other.
resume_sourcing_automation
ChatGPTResumes a paused Sourcing automation where it stopped; contacting suppliers continues. It never agrees payment, samples or terms (those wait for the owner as approvals), and the 25% margin / 150% ROI gate can be raised, never lowered.
run_patent_check
ChatGPTRuns and saves a USPTO patent check on a research project: up to 5 phrases describing the product against Patent Public Search (title / abstract / claims). Queries the public USPTO patent site; free, no key; 12 checks a minute. Read the verdicts afterwards with get_ip_checks.
run_trademark_check
ChatGPTRuns and saves a USPTO trademark check on a research project: brand names and / or product keywords against the trademark index in the given Nice classes (e.g. 21 kitchen, 20 furniture, 28 toys, 16 stationery, 35 retail). Queries the public USPTO trademark site; free, no key; 12 checks a minute. Read the verdicts afterwards with get_ip_checks.
search_again_sourcing_automation
ChatGPTAsks a running Sourcing automation for another supplier search on Alibaba while it is finding suppliers; new suppliers only, never ones already contacted. It never agrees payment, samples or terms (those wait for the owner as approvals), and the 25% margin / 150% ROI gate can be raised, never lowered.
search_brain
ChatGPTFind company brain notes whose path contains the given words (e.g. "supplier", "gates", "roles", "resale"). Returns paths to open with get_brain_note. Without a query it lists every note.
search_product_ideas
ChatGPTFinds candidate products with Product Ideation: Keepa's product database filtered by a Sotrics criteria preset, then Sotrics' fee and revenue checks. preset: sotrics-low-ticket | sotrics-basic | sotrics-standard | new-seller | low-competition | premium (sotrics-low-ticket = under $15, 10+ sales a day, up to 1,000 reviews; sotrics-basic / sotrics-standard = the $18-25+ products). keywords = words the title must contain (the niche, e.g. "stickers"). Returns up to 50 products per page with price, monthly sales, revenue, reviews, rating, BSR and fee %. Spends Keepa tokens (daily limit per account). Next: add_ideas_to_pipeline for the ASINs worth validating.
set_research_status
ChatGPTChange the pipeline stage of a product research project (id from get_product_pipeline), or archive / unarchive it. Stages: new, researching, market_validated, keyword_validated, review_analyzed, financial_validation, approved, sourcing, sample_testing, ready_to_launch, launched, rejected, archived.
set_tracked_keywords_active
ChatGPTPause (active=false) or resume (active=true) rank-tracker keywords by id (ids from get_rank_tracker). Paused keywords keep their history. Nothing is deleted.
start_research_run
ChatGPTQueue Research Automation runs (one per keyword, max 10 per call; US by default): SERP capture, enrichment, basic check, AI segmentation, keyword analysis, offers, landed cost, then ONE decision (decide_research_run), trademark / patent and the hand-off to sourcing. Spends provider credit; real accounts only. Progress: get_research_runs / get_research_run.
start_sourcing_automation
ChatGPTStarts the Sourcing automation of a Supplier Tracker product: finds Trade Assurance, verified Alibaba suppliers in waves and messages them through the owner's own Alibaba account and worker until 3 suppliers pass the gate. Real messages go to real suppliers. Needs SOURCING_AUTOMATION=on, a sender account and the worker on the owner's computer. It never agrees payment, samples or terms (those wait for the owner as approvals), and the 25% margin / 150% ROI gate can be raised, never lowered.
stop_sourcing_automation
ChatGPTStops the Sourcing automation of a product for good: no more suppliers are contacted and the wave progress is lost. Quotes already recorded stay in the Supplier Tracker. It never agrees payment, samples or terms (those wait for the owner as approvals), and the 25% margin / 150% ROI gate can be raised, never lowered.
update_alert
ChatGPTWork on one Sotrics alert (id from get_alerts): resolve it (optional note), reopen it, mark it under investigation, snooze it for some hours, or add a note. Changes only the alert in Sotrics, nothing on Amazon.
update_product_costs
ChatGPTSave landed costs per SKU (product cost, freight, shipping/prep to Amazon, other), in the marketplace currency. Fields left out keep what is stored. Every profit number (CM1-CM3, P&L, health checkup) uses these costs, so confirm the values with the person first. At most 50 SKUs per call.
update_supplier_quote
ChatGPTChanges one supplier quote: its status in the sourcing flow (CONTACTED -> QUOTED -> SHORTLISTED -> SAMPLE_ORDERED -> SAMPLE_APPROVED -> SELECTED / REJECTED, with rejectReason) and any quote fields the supplier updated. Selecting a supplier or ordering a sample is the owner's decision (it becomes a work order).