com.webotee/amazon-product-research-mcp

Find UNDER-SERVED NICHES — real SUBCATEGORIES with genuine demand but room to compete, ranked by private-label winnability from 2-year marketplace data. Returns CATEGORIES / sub-categories (e.g. 'Wireless Earbuds', 'Cable Organizers'), NEVER brands. This is the RIGHT tool for ANY niche-discovery question: 'under-served niches', 'niches in <category>', 'find a niche to enter', 'what niche should I sell in', 'underserved categories', 'gaps in <department>', 'where's the opportunity in <category>'. When the user names a department or category (e.g. 'electronics', 'home & kitchen'), pass it as category_name to scope the niches to that area. Do NOT use category_undercompeted_brands or find_undercompeted_brands for niche questions — those return BRANDS, not niches.

Recommend ungate / arbitrage / avoid for an ASIN, with a 3-bullet rationale citing named metrics (gating_score, amz_retail_dominance_pct, fba_pct, brand_posture). Also returns the ASIN's product brand, title and price (or price range) plus fulfillment (FBA/FBM/Amazon). Use for 'should I try to ungate this?' / 'how should I price this?'.

Check which of the 11 Amazon marketplaces a keyword was observed in Amazon's own autocomplete suggestions, with first/last observed dates and an example current suggestion-list position per marketplace. Use when a seller asks 'does anyone type X on Amazon Germany/Japan/…', compares keyword presence across countries, or plans a marketplace expansion (pair with brand_xmarket / operator_xmarket_presence). Exact-keyword match — not volumes, not rankings. Amazon marketplaces only.

Keywords observed in Amazon's own autocomplete suggestions for a seed term, per marketplace: the current suggestion list(s) for the seed's prefix (each term with its position 1-10 within that list) plus related observed vocabulary starting with the seed, with the marketplaces each term was observed in. Use for listing/backend keyword language, 'what do buyers type for X', or seeding niche/product research with real buyer phrases. No volume figures and no organic-ranking data — observed suggestion vocabulary only. Amazon marketplaces only.

Look up products by retail barcode: UPC-12, EAN-13 or GTIN-14. Use when the user gives a numeric product barcode (from a shelf tag, an invoice, a supplier price list, a wholesale catalog) and wants to know which Amazon or Walmart listing it maps to — e.g. 'what is UPC 050875825598 on Amazon', 'match these barcodes to ASINs'. Returns the matched products (best match first — priced and recently observed rows lead) with ASIN, brand, title, price, rating and image. One barcode per call.

Show active MAP (Minimum Advertised Price) violations for products in the workspace. Use when the user asks 'MAP violations', 'who is selling below MAP', 'price violations today', 'are there any MAP breaches', or any MAP-enforcement question.

Who's competing with the seller on their OWN products. Pass an ASIN from your catalog to see the full competitor list for that listing (who wins the buy-box, FBA/FBM, observed price, how long they've been on it) plus your own buy-box / undercut status. Omit the ASIN to get your most-contested products (most competing sellers / where you're being undercut). Requires a connected store (Starter+). Use for 'who am I competing with', 'am I losing the buy-box', 'who else sells B0...', 'where am I under pressure'.

What to source NEXT, adjacent to what the seller already sells. Seeds Webotee's undercompeted-brand and underserved-niche engines from the seller's OWN connected catalog (their dominant brands and categories), excluding brands they already carry. Requires a connected store (Starter+). Use for 'what should I source next', 'expand my catalog', 'adjacent opportunities', 'what else could I sell'.

The seller's OWN connected Amazon catalog fused with Webotee market intelligence. Each product shows EXACT account data (your price, your inventory, FBA/FBM, listing status) alongside OBSERVED/ESTIMATED market data (sourcing score, the observed market buy-box price, how many sellers are on the listing, estimated 30-day units, cross-marketplace spread) and a price_vs_market read (below_market / at_market / above_market). Requires a connected store (Starter+). Use when the seller asks about their own products, 'how am I priced vs the market', 'which of my products are under pressure', or 'show my catalog'.

Store-level summary of the seller's connected Amazon catalog: number of products tracked, distinct brands, FBA vs FBM split, total on-hand inventory units (EXACT, from the account), average Webotee sourcing score, how many products are currently undercut, and how many have a cross-marketplace (Walmart) opportunity. Requires a connected store (Starter+). Use for 'how is my store doing', 'summarise my catalog', or a dashboard overview.

Given one or more operator (seller) names, return the brands they carry ranked by seller competition level. Use when the user asks 'which brands sold by these operators have the least competition', 'out of these operators, show me brands with fewest sellers', 'low-competition brands for operator X', or any follow-up that chains operator names to brand-level competition metrics.

For ONE operator (seller), find the ASINs where it's losing the buy-box — it carries the listing but holds less than `max_buybox_share_pct` of the buy-box — ranked by estimated 30-day revenue at risk. Each leak returns who's winning the box, the operator's vs the winner's sanitized price and the gap, how many sellers are on the listing, a `situation` (price_gap | tied_price | amazon_in_box | reseller_swarm), a `suggested_action` (reprice_to_win | enable_fba | defend_match_only | evaluate_or_exit), and a `suggested_target_price` when the box is winnable on price. Totals include the recoverable price-gap-only revenue. Use when the user asks 'where is seller X losing the buy-box', 'X's buy-box leaks', 'which ASINs should X reprice', 'where is X leaving money on the table', or 'what's X's revenue at risk'. Sortable by revenue-at-risk (default), price gap, buy-box share, or listing revenue.

Show which product categories an operator dominates. Returns categories ranked by the operator's brand count and observed buybox days in each. Use when the user asks 'what categories does this seller focus on', 'operator category breakdown', 'where does Amazon Warehouse dominate', or any category-scoped operator question.

Label what kind of operator a seller is — amazon, brand-direct, likely-authorized-retailer, arbitrage, or reseller — from our cross-brand operator signals (how many brands they span, their fulfilment mix, their primary brand). Use when the user asks 'what kind of seller is this', 'is this an authorized retailer or an arbitrage seller', 'classify this operator'. Heuristic label, not a legal determination. Amazon US/UK.

ASIN-level buy-box head-to-head between two operators (sellers). For the ASINs BOTH carry (contested), returns who holds the buy-box (winner), each operator's buy-box share, their sanitized average prices and the price gap, how many OTHER sellers are on the listing, whether Amazon holds the box, and a `situation` tag (amazon_in_box | reseller_swarm | tied_price | price_gap) — plus operator_a's estimated 30-day revenue-at-risk. Also returns each side's WHITESPACE: high-revenue ASINs that operator wins which the OTHER doesn't carry (expansion targets). Use when the user asks 'compare seller A vs seller B', 'where do these two sellers overlap / compete', 'who wins the buy-box between them', 'show me the head-to-head', or 'what does A sell that B doesn't'. Sortable by revenue-at-risk (default), price gap, or share gap.

Operator concentration (HHI) across a set of ASINs or a brand's catalog — reveals when the same operator quietly controls many of the 'independent' listings. Returns the HHI, a concentration label, and the top operators with how many listings each dominates. Use for private-label fragmentation checks: 'how concentrated is this niche', 'who controls this brand's listings', 'is one operator running most of these'. Amazon US/UK. Optional filters (all default to no filter): min/max hhi (gates the whole analysis by its concentration index); on the top_operators list — operator (exact, case-insensitive) + operator_contains, min/max share_pct, min/max listings_controlled.

Show brands an operator recently stopped selling (churn signal). Use when the user asks 'what brands did this seller drop', 'operator churn', 'brands lost by X', or any question about an operator shrinking their catalog.

Show brands an operator recently started selling. Use when the user asks 'what new brands did this seller pick up', 'operator new brands', 'what is Amazon Warehouse selling now that it wasn't before', or any question about an operator expanding their catalog.

Operators newly OBSERVED on a brand in a recent window — counted at BRAND level: a seller's FIRST observation anywhere across the brand's ASINs falls in the window. This is coverage-robust (a long-present seller was almost certainly seen on some ASIN earlier, so they correctly drop out) — a trustworthy directional 'new on the brand' count, not the inflated per-ASIN number. Still first-OBSERVED, not provably first-to-market. Returns each operator with first_observed, how many of the brand's ASINs we've seen them on, and whether still active. Use for 'who's new on <brand>', 'who's newly showing up on my brand', 'recent sellers on <brand>'. Amazon US/UK. since_days already bounds first_observed below; optional filters (all default to no filter): operator (exact, case-insensitive) + operator_contains, first_observed_from/_to, min/max n_asins_on_brand, min/max total_days_seen, min/max observed_buybox_days, still_active (true/false).

The arbitrage sourcing feed: given an ASIN (or a seller/operator name), find the OTHER ASINs the operators on that listing also resell, ranked by sourcing fit (our composite sourcing score) then estimated 30-day sales. De-duped against the ASINs the user already tracks, so it surfaces NEW candidates. Use when the user says 'what else does this seller carry', 'find more like this from the same operators', 'arbitrage leads from this ASIN's sellers'. Each candidate ASIN also carries product brand, title, and price (or price range) plus its fulfillment (FBA/FBM/AMZ + amz/fba pct). Amazon US/UK. Optional filters (all default to no filter): min/max sourcing_score, est_units_30d, buybox_avg_price; resold_by_contains; the shared product/fulfillment block (product_brand, product_title_contains, min/max_price, fulfillment_in, amz/fba pct).

Show the ASINs an operator wins the buybox on most often, ranked by ESTIMATED 30-day sales by default. For each ASIN it returns the operator's estimated units sold and revenue in the last 30 days (est_units_30d, est_revenue_30d — the product's sales estimate weighted by the operator's buy-box share) plus the operator's BUYBOX SHARE (percent of observed days it held the buybox; normalized, not raw days). Sortable by est_sales (default), observed buybox days won, price, or days seen. Use when the user asks 'what ASINs does this seller win on', 'top ASINs for operator X', 'what does this seller sell the most of', 'best products for this seller', or any ASIN-level operator drill-down. Each ASIN also carries product brand, title, and price (or price range) plus its fulfillment (FBA/FBM/AMZ + amz/fba pct).

Show the brands an operator sells the most of, ranked by ESTIMATED 30-day sales by default. For each brand the operator carries it returns the operator's estimated units sold and revenue in the last 30 days (est_units_30d, est_revenue_30d — the estimated sales of the ASINs the operator wins for that brand, weighted by its buy-box share), the number of the brand's ASINs the operator wins, and observed buybox days. Use when the user asks 'what brands does this seller sell the most of', 'top brands for operator X', 'which brands make this seller the most money', or any brand-level operator drill-down by sales. For brand competition (fewest sellers) instead of sales, use operator_brands_by_competition.

Check whether an operator sells on Amazon US, Amazon UK, and/or Walmart. Returns per-marketplace brand count, ASIN count, and observed buybox days. Use when the user asks 'does this seller sell on Walmart too', 'cross-marketplace presence', 'is this operator on Amazon UK', or any multi-marketplace operator question.

Save a reusable per-model research workflow (a 'playbook') the user can re-run or schedule. Provide a template_key (one of: brand_watch, new_brand_radar, replenishment_watch, arbitrage_feed, defend_my_niche, find_my_next_niche, brand_defense_daily, expansion_radar, dropship_watch, spread_hunter, map_sweep, operator_network_expose, gating_risk_guardian) with its scope, OR custom steps. scope holds the inputs every step shares (e.g. {"brand":"Nike"} or an ASIN). Use when the user says 'save this as a weekly check', 'make a playbook for ...', 'automate this research'.

List the workspace's saved playbooks (name, schedule, last run) and the built-in templates available. Use when the user asks 'what playbooks do I have', 'show my saved workflows', 'what automations are set up'.

Run a saved playbook right now and return its digest (also saved to the in-app inbox). Use when the user says 'run my <name> playbook', 'check my brand watch now', 'run that workflow'.

Set how often a saved playbook runs automatically (manual, daily, or weekly). Use when the user says 'run this weekly', 'schedule my playbook daily', 'stop the automatic runs' (manual).

The observed page-content record for one Amazon ASIN: description (with its source — prose vs 'About this item' bullets vs brand A+ content), feature bullets, breadcrumb category path, image-gallery URLs, variation count + parent ASIN, stock state, and first/last-observed timestamps. Set include_reviews=true to append the featured customer reviews shown on the page (author, rating, title, trimmed text, verbatim date, verified flag, helpful votes). Use when the user asks what a product's page says or shows — its description, bullets, images, categories, variations — or wants the reviews on a listing. Amazon marketplaces only.

Risk + protection assessment for an ASIN or brand. Returns composite risk score (0-100), recent MAP violation events (≤10, each with the offending ASIN's product brand, title, price or price range and fulfillment FBA/FBM/Amazon), unauthorized seller list (≤10), and 1-3 recommended actions. Use for 'flag risk events on my brand' or 'is this ASIN risky?' style prompts.

Free-text keyword search over Amazon product titles + descriptions. Use when the user names a PRODUCT TYPE or keywords (e.g. 'stainless steel water bottle', 'cat litter box', 'magnetic phone mount') rather than a specific brand, category, or ASIN. Returns the top matching products ranked by relevance with brand, price, 30-day demand, fulfillment (FBA/Amazon/FBM) and rating. Optional filters narrow the result by any returned field: product_brand, min/max_price, fulfillment_in (FBA/FBM/AMZ), min/max demand, and min/max rating.

Record the user's cost basis (landed unit cost) for an ASIN so later checks can flag when the buy box drops below their cost and compute true net spread; the saved/cleared confirmation also echoes the ASIN's product brand, title and price (or price range) and its fulfillment (FBA/FBM/AMZ). Use when the user says 'my cost on B0... is $12', 'set my cost basis', 'I pay $X for this'. One ASIN per call; pass cost 0 to clear.

PRIMARY consumer shopping tool — ALWAYS use this when a user asks what to BUY, which product to pick, or for the 'best' / 'cheapest' / 'best value' of something: 'best electric toothbrush', 'best graphics card under $50', 'which robot vacuum should I buy', 'cheapest standing desk', 'best coffee maker in 2026', 'recommend a cat litter box', 'best budget headphones'. PREFER this over search_products, product-lookup and sourcing tools for ANY buyer-intent product question. Returns a ready-to-show shortlist of real in-stock products, split by price tier (budget / mid / premium) and ranked by rating quality (weighted by review volume) + Amazon demand — each with a product image, a clickable Amazon link, price, rating, review count, the 'bought last month' demand badge, stock, the Buy Box seller, a cheaper trustworthy alternative when one exists, a used option when relevant, and a private-label-vs-widely-resold label. Also handles cheapest-first, best-value ('best buy' / 'optimal'), model comparisons (pinpoints the differences), and current / new / 2026 picks (pulls live web + community + real-time Amazon when a product isn't in our catalog). Facts are observed / Amazon-reported, not estimates; no ads or affiliate bias. (Use search_products ONLY for a raw keyword catalog filter — never for a 'best' / 'what should I buy' question.)

Find sellers (operators) expanding into the most NEW brands in a recent window. Use when the user asks 'operators expanding into new brands', 'sellers growing fastest by brand count', 'who is moving into new brands this month', or any cross-cutting operator question without a specific seller named.

Find top sourcing-pick ASINs across the entire catalog with optional filters: max retail price, min retail price, category, brand, exclude gated, exclude Amazon private label. Use when the user asks 'top sourcing picks', 'best ASINs to source', 'ASINs under $X with rising demand', 'fastest growing ASINs', or any cross-cutting question where they have NOT named a specific entity yet. Each pick carries product brand, title and price (or price range) plus fulfillment (FBA/FBM/Amazon) alongside the sourcing scores.

Find top brands by 90-day unit velocity (brand_velocity_90d_units_day). Optional filters: scope (tracked = user's watchlist, universe = all), category, minimum velocity, exclude Amazon private label, exclude gated. Use when the user asks 'fastest selling brands', 'top velocity brands', 'brands I track by velocity', 'what brands move the most units?', or 'best selling brands in [category]'. When the user says 'my brands' or 'brands I track', set scope=tracked.

List the sellers winning a brand's buy box — the resellers and arbitrage operators you're up against — each classified (authorized-retailer / arbitrage / Amazon / brand-direct / reseller). If you've saved an authorized list (authorized_seller_set) it instead flags the UNAUTHORIZED sellers. Use when the user asks 'which operators dominate the buy box on <brand>', 'who else is selling my brand', 'unauthorized sellers on Nike', 'find rogue sellers', or any brand buy-box / protection question.

Create or add to a saved tracking list the user can monitor over time. list_type is one of asin, brand, seller, niche; name is the user's label for the list; items are the identifiers to track (ASINs, brand names, seller names, or niche keys). Captures a baseline of the current observed state so a later 'what changed' check can show new sellers and score moves. Use when the user says 'track these ASINs', 'add Nike to my brand watchlist', 'start monitoring ...'.

Resolve a natural-language watchlist request into concrete ASINs the user can add. Use when the user says 'watch this brand', 'alert me when X loses a seller', 'add Nike to my watchlist', or any watchlist-creation intent. Returns matching ASINs with current scores plus product identity (brand, title, price or price range) and fulfillment (FBA/FBM/AMZ with amz/fba share) so the user can confirm which to add.