MCP App Store
by KeenEthics
EzSeller icon

EzSeller MCP Server for ChatGPT

By SMAT Designs

Overview

Use EzSeller when a question depends on the seller's own authorized Amazon Seller Central evidence: listing suppression or requirements, FBA inventory discrepancies, Sales & Traffic and search-performance changes, Seller Account Health, estimated FBA storage fees, promotion performance, Amazon financial events, selling-fee estimates, and reimbursement audits. EzSeller keeps seller, marketplace, product identity, observed evidence, derived conclusions, pending provider work, and unknowns distinct. The public ChatGPT connection is read-only for Amazon state: it does not modify Amazon Seller Central listings, file reimbursement claims, control Amazon Ads campaigns, or invent missing external costs needed for complete landed-profit calculations.

How to connect

  1. ChatGPT: Open the EzSeller listing in the ChatGPT app directory and choose Connect. ChatGPT listing EzSeller exposes 114 tools here.

Tools

amazon_ad_spend_request

ChatGPT
Request tenant-bound Sponsored Products spend reports by campaign, ad group, or ASIN. Unsupported ad products are identified explicitly. Reports are async: poll amazon_ad_spend_results with each returned reportId.

amazon_ad_spend_results

ChatGPT
Poll tenant-bound Amazon Advertising reports and read observed spend and supported 14-day attributed metrics by campaign, ad group, or ASIN. Results retain native currency, source periods, evidence timestamps and missing-data limits.

amazon_ads_budget_pacing

ChatGPT
Audits campaign daily budgets and hourly spend pacing to flag early budget exhaustion, under-pacing campaigns, and profitable campaigns capped by daily budget constraints.

amazon_ads_manager_profiles

ChatGPT
Enumerate the advertising profiles reachable under one Ads Manager link's agency authorization (GET /v2/profiles fanned out over the NA, EU and FE Ads endpoints) and attribute each profile: seller profiles bind to the client org whose in-scope connection carries the same seller id (exactly one match required); agency-type profiles are the manager account's own; anything else is returned as unattributed with a reason, never pooled into a client. Requires linkId plus exactly one selector: clientOrgId (only that client's profiles) or portfolioScope:true (all in-scope groups). A missing or revoked agency authorization returns the fail-closed amazon_ads_authorization_required payload. One region failing does not sink the others; partial data is labelled.

amazon_ads_portfolio_summary

ChatGPT
Cross-client Ads rollup for one Ads Manager link (portfolioScope:true required): per client org — profile count, per-currency profile counts and total daily budgets (summed only within a currency, never blended), and marketplace ids — with the manager account's own agency-level profiles summarized separately and unattributed profiles counted with reasons. Built from the same profile enumeration as amazon_ads_manager_profiles; spend reporting stays with the per-tenant Ads tools. Carries an observedAt freshness stamp.

amazon_ads_search_term_harvest

ChatGPT
Analyzes customer search terms from auto/broad campaigns to identify profitable queries to promote to Exact Match (Search Term Harvester) and non-converting high-spend queries to negate (Bleeder Radar). Generates pre-mutation snapshots for approval before applying keyword changes.

amazon_ads_tacos_intelligence

ChatGPT
Calculates storewide True TACoS (Total Advertising Cost of Sales), organic-to-paid ratio, and organic revenue multiplier. Reconciles durable Sales & Traffic ordered revenue against advertising spend evidence to diagnose true ad dependency and organic halo effect.

amazon_agency_client_report

ChatGPT
Use this when an agency user asks for one client's report. Builds the report from that client's own account data (account health, connection health, reimbursement queue, alerts; or, when period is supplied, composed weekly/monthly/custom period sections including headline, listing exceptions, money history, seller pulse, and reimbursement case drafts) and renders it under the client's own branding. Requires an explicit clientOrgId — there is no default and no portfolio mode — and the caller must hold an assignment covering that client for the report function; any other client is refused with client_scope_denied before any data is read.

amazon_analytics_runs

ChatGPT
List tenant-bound analytics backfill/schedule runs with provider query/report IDs so accepted operations can be polled with the matching results tool.

amazon_analytics_schedule_create

ChatGPT
Create a tenant-bound recurring Data Kiosk, SQP, or SCP collection schedule and enqueue only complete historical backfill periods.

amazon_analytics_schedules

ChatGPT
List this tenant and seller connection’s analytics schedules, run counts, and ambiguous dispatch counts.

amazon_answer_seller_question

ChatGPT
Ask a real Amazon seller question: why sales changed, what fees Amazon charged, whether an order shipped, whether A+ Content is published, what mobile-app traffic was, what buyers complained about, or why inventory is unavailable. Set resolveEvidence:true to retrieve a scoped source-backed read. For Performix historical month comparisons, also supply priorMonth and currentMonth as YYYY-MM. Agency SKU listing and FBA reads require exactly one assigned connectionId or clientOrgId; no default seller is selected, and other agency source requests remain unavailable. A plan alone is not an answer. Never infer bank deposits, actual profit, ad incrementality, mobile sales or buyer identities; never issue a write from this tool.

amazon_aplus_asin_relations

ChatGPT
Read ASIN relationships, badges and content linkage for one seller-authorized A+ document. Use Amazon publish records to prove publication independently.

amazon_aplus_content_state

ChatGPT
Read Amazon A+ publish records for an ASIN. Establishes publish-record evidence only; it does not score content quality.

amazon_aplus_document

ChatGPT
Read a seller-authorized A+ Content document, including its content modules and metadata, by Amazon content reference key. Never interprets document existence as publication.

amazon_aplus_documents

ChatGPT
List A+ Content documents owned by the authorized seller in the selected marketplace. Shows provider metadata and pagination, not publication or performance claims.

amazon_aplus_inventory

ChatGPT
Inventory a bounded seller-authorized A+ corpus and classify Premium, Standard, Brand Story, or unknown only from Amazon document badges, content type, and observed module types. Reads documents only; never infers Premium eligibility from document existence.

amazon_aplus_media

ChatGPT
Read one seller-authorized Amazon A+ media asset and its related media associations. Marketplace selection validates seller authority; media identity comes from Amazon and this tool never creates or changes media.

amazon_aplus_validate_asin_relations

ChatGPT
Validate an Amazon A+ Content document against a bounded ASIN set via Amazon content ASIN validations. Read-only; mutates nothing at Amazon.

amazon_approve_buyer_outreach

ChatGPT
Human owner/admin approval of an exact Amazon outreach preview. Approval alone sends nothing; execution is separately guarded.

amazon_brand_store_insights

ChatGPT
Read actual Amazon Ads Store insights for a Store proved to belong to the connected advertiser; supports Store/date/page/traffic-source dimensions for page views, visits and Amazon-attributed sales. Read only, one metric per request.

amazon_brand_stores

ChatGPT
List the authenticated advertiser account's Brand Stores and their Amazon brand entity identifiers through the separate Amazon Advertising API. Does not use SP-API sales data as Store analytics.

amazon_case_lifecycle_bind

ChatGPT
Bind a seller-filed Amazon Case ID to a case lifecycle record. Fails closed without an authoritative citation or on conflict.

amazon_case_lifecycle_create

ChatGPT
Create a drafted Amazon Seller Support case lifecycle record from an A16 draft shape or native_manual core. Re-verifies cross-agency claim safety for reimbursement claims.

amazon_case_lifecycle_follow_up_draft

ChatGPT
Draft a follow-up packet for an existing Seller Support case citing only stored record facts and new evidence. Draft only — never sends.

amazon_case_lifecycle_follow_up_sent

ChatGPT
Record that the seller has sent a follow-up in Seller Central. Status is marked sent_unverified until confirmed via thread observation.

amazon_case_lifecycle_get

ChatGPT
Retrieve a single Seller Support case lifecycle record by internal lifecycle ID or Amazon Case ID. Output always includes provenance, lifecycleStatus, amazonSurfaceStatus, and replyObserved.

amazon_case_lifecycle_import_history

ChatGPT
Idempotently import historical Amazon Seller Support cases from lobby observations. Never overwrites existing transitions; reconciles claimed lobby total against enumerated rows with honesty reporting.

amazon_case_lifecycle_list

ChatGPT
List Seller Support case lifecycle records for the current connection. Every record carries provenance, lifecycleStatus, amazonSurfaceStatus, and replyObserved.

amazon_case_lifecycle_observe

ChatGPT
Append an observed surface status and/or Amazon reply to a case lifecycle record. Updates state machine based on authoritative citation.

amazon_case_lifecycle_resolve

ChatGPT
Record an authoritative resolution on a filed Amazon Seller Support case. Transitions local state to resolved.

amazon_case_packet_draft

ChatGPT
Use this when the user asks to draft an Amazon Seller Support case packet or Plan of Action (POA) for listing removals, NWD brand-role flags (errors 5465/6045), account health warnings, product safety appeals, or reimbursement packets from cited record facts. Every statement is grounded in cited evidence; missing facts are labeled UNKNOWN, never invented. DRAFT ONLY — never submits; the seller reviews and files manually in Seller Central.

amazon_connection_health

ChatGPT
Live diagnose Amazon onboarding health: DB row + LWA refresh + marketplace participation + freshness.

amazon_connection_status

ChatGPT
Read the tenant-bound Amazon connection row from Neon (DB only). Prefer amazon_connection_health for live LWA proof.

amazon_customer_feedback

ChatGPT
Read Amazon customer-feedback topics and trends for an ASIN, plus category review/return topics when Amazon exposes them. Surfaces per-topic mention counts in reviewSignal. Provider observations are evidence, not causal diagnosis; Amazon exposes no total review count or review-velocity series here.

amazon_data_kiosk_sales_traffic_next_page

ChatGPT
Request the next page of a completed tenant-bound Data Kiosk query. EzSeller reuses Amazon’s authoritative query and opaque token; callers cannot supply GraphQL or tokens.

amazon_data_kiosk_sales_traffic_request

ChatGPT
Submit a tenant-bound, schema-pinned Amazon Data Kiosk Sales & Traffic query. Arbitrary GraphQL is not accepted.

amazon_data_kiosk_sales_traffic_results

ChatGPT
Poll one tenant-bound Data Kiosk Sales & Traffic query and return bounded JSONL records after Amazon finishes.

amazon_diagnose_listing

ChatGPT
Use this when the user asks why an Amazon listing or seller SKU is suppressed, inactive, not buyable, undiscoverable, or showing listing issues. Reads current seller-scoped listing status and Amazon issue evidence for one SKU. Do not use for market-wide product research or competitor listings.

amazon_diagnose_listing_quality

ChatGPT
Diagnose external IDQ/NSI+/support catalog-quality findings against tenant-bound live Amazon seller listings and Catalog Items. Returns confirmed/contradicted/not-established evidence and routes supported repairs into the guarded listing workflow.

amazon_discover_listing_workflow

ChatGPT
Search catalog by identifier and classify new-offer vs restricted vs ambiguous. Read only.

amazon_estimate_fees

ChatGPT
Use this when the user asks what Amazon would charge in selling fees for one SKU or ASIN at a proposed price. Returns referral, fulfillment, and other fee-estimate details plus the total from the Amazon fee-estimate interface. This is an estimate, not settled fees, complete landed cost, or guaranteed profit.

amazon_explain_inventory

ChatGPT
Use this when the user asks where FBA units went, why Amazon inventory differs from expectations, or what Amazon currently observes for one seller SKU. Returns seller-scoped FBA inventory evidence and preserves missing quantities as not observed instead of inventing zero. Do not use for demand forecasting or warehouse-planning projections.

amazon_explain_money

ChatGPT
Use this when the user asks what Amazon charged, credited, reimbursed, or recorded financially for an order ID, financial-event group, or posted date window. Returns item-level components and reconciliation evidence for the observed Amazon financial events. Do not present it as bank payout proof or complete landed profit when COGS, freight, ads, storage, or other external costs are not observed.

amazon_fba_shipment_status

ChatGPT
Read and reconcile the current SP-API inbound plan status and shipments for a durable FBA inbound shipment operation. Does not replay Amazon writes.

amazon_fee_schedule_compare

ChatGPT
Use this when the user asks what Amazon's announced 2027 fee changes do to one SKU's fulfillment fee, whether a sub-$10 item loses Low-Price FBA rates, or what a Buy X Get Y promotion will cost under the new promotion fee. Prices one unit under both schedules from the caller-supplied current fee plus Amazon's published 2027 band deltas, and prices the BXGY promotion fee ($5 + 2% capped at $2,000, from Jan 15 2027). Published schedules only — not settled fees and not a substitute for Amazon's Fee Change Comparison dashboard.

amazon_flat_file_submission_status

ChatGPT
Poll and reconcile a durable Amazon flat-file submission operation. Read-only; never replays Amazon writes. On DONE downloads and parses the processing report with per-row results and counts.

amazon_get_order

ChatGPT
Read one authoritative order from Amazon Orders API 2026-01-01 by order ID using the same safe seller-operational projection as amazon_search_orders.

amazon_inventory_forecast

ChatGPT
Extrapolate evidence-bounded demand, days of cover, reorder point, order-by date, and suggested restock for one seller SKU from live Amazon FBA inventory plus persisted Sales & Traffic velocity history. Withholds projections when sales history is stale or missing.

amazon_inventory_health_request

ChatGPT
Request the official Amazon FBA Inventory Health report for the seller-authorized marketplace. Report creation is an external control action but does not mutate seller catalog, inventory, orders, or pricing.

amazon_inventory_health_results

ChatGPT
Read a requested Amazon FBA Inventory Health report. Returns bounded seller-operational rows from the official Amazon report; customer comments and direct contact/address fields are omitted from MCP output.

amazon_inventory_risk

ChatGPT
Calculate evidence-bounded days cover and stockout risk for one seller SKU using current Amazon FBA inventory plus the newest persisted Sales & Traffic observation for the same ASIN/SKU. Returns unknown rather than forecasting when inventory, sales velocity, or recency evidence is insufficient.

amazon_list_pending_approvals

ChatGPT
Use this when an agency user or agent asks which listing writes are waiting for approval. Returns the tenant's pending approvals scoped to the clients the caller's assignments cover, each row attributed to the client it was planned against and stamped with the requesting principal, plus a canApprove flag and a stale-binding flag when the connection has been reassigned to another client. Read-only: approving stays a human act enforced by the operation ledger. Requires portfolioScope:true; a single-connection target is refused.

amazon_listing_live_read

ChatGPT
Read live Listings Items plus PTD and restrictions for one SKU/ASIN/product type.

amazon_listing_restrictions

ChatGPT
Read listing restrictions for one ASIN on the connected marketplace.

amazon_low_inventory_fee_check

ChatGPT
Use this when the user asks whether the low-inventory-level fee will apply to a SKU under Amazon's revised test (Feb 1 2027), or why a low-inventory charge is not visible yet. Evaluates the published two-sided 28-day test: the fee applies only when days of supply are below 28 on BOTH the trailing 28 days of shipments AND the following 28 days. When the following window has not elapsed, the result says the charge is unknowable yet — by Amazon's design it surfaces ~4 weeks late. Applicability only; per-unit fee rates are unchanged since Apr 2024 and are not modeled.

amazon_plan_buyer_message

ChatGPT
Read live order eligibility and preview the exact Amazon POST body. Seller-supplied text is Expert-preview-only; no sends occur in plan.

amazon_plan_review_request

ChatGPT
Read live order eligibility and preview the exact Amazon POST body. Seller-supplied text is Expert-preview-only; no sends occur in plan.

amazon_portfolio_account_health

ChatGPT
Use this when an agency user asks for an account-health view across clients. This is a labeled composition of connection-health gates plus LISTINGS_ITEM_ISSUES_CHANGE notification signals per connection — EzSeller has no dedicated SP-API account-health read, and this tool never presents the composition as an Amazon account-health score. Requires portfolioScope:true.

amazon_portfolio_alerts

ChatGPT
Use this when an agency user asks what needs attention across their client accounts. Returns recorded Amazon notification signals (report processing finished, listing issue changes, and any future signal type passed through opaquely) for the caller's in-scope connections only, each with client attribution. Requires portfolioScope:true.

amazon_portfolio_health

ChatGPT
Use this when an agency user asks how all of their client seller accounts are doing at once. Returns per-connection connection-health gates with client attribution, gate status counts, and a coded error row for any connection that cannot be read (for example a revoked token) instead of failing the whole rollup. Requires portfolioScope:true; a single-connection target is refused.

amazon_portfolio_pulse

ChatGPT
Use this when an agency user asks how sales are doing across all of their clients. Returns the latest materialized Seller Pulse snapshot per in-scope connection with client attribution, plus sales totals grouped strictly by currency and comparison window (never blended across currencies or windows). Requires portfolioScope:true.

amazon_portfolio_reimbursement_queue

ChatGPT
Use this when an agency user asks what Amazon still owes across their client accounts. Returns the cross-client FBA/reimbursement work queue with per-client attribution, amounts sorted by amount at risk within each currency, and exact per-currency totals (currencies are never blended). Requires portfolioScope:true.

amazon_product_research

ChatGPT
Search Amazon catalog by keyword and return seller-authorized marketplace product/classification/sales-rank evidence. Rank is not a guarantee of demand, margin, or future sales.

amazon_product_type_definition

ChatGPT
Read the Product Type Definition summary for one product type.

amazon_promotion_performance_request

ChatGPT
Request the official Amazon Promotion Performance report for the seller-authorized marketplace. Report creation is an external control action but does not mutate seller catalog, inventory, orders, or pricing.

amazon_promotion_performance_results

ChatGPT
Read a requested Amazon Promotion Performance report. Returns bounded seller-operational rows from the official Amazon report; customer comments and direct contact/address fields are omitted from MCP output.

amazon_promotions

ChatGPT
Read Amazon promotion state for coupons, deals, price discounts, and basket-building promotions. Running status is not evidence of profitability.

amazon_reconcile_catalog_quality

ChatGPT
Reconcile external IDQ/NSI+/support catalog-quality findings against tenant-bound live Amazon seller listings and Catalog Items. Returns confirmed/contradicted/not-established evidence and routes supported repairs into the guarded listing workflow.

amazon_reconcile_listing_operation

ChatGPT
Reconcile a durable listing create, repair, or canonical identity operation. Does not replay Amazon writes.

amazon_reconcile_settlements

ChatGPT
Check whether Amazon settlement payouts for the authenticated seller match the totals Amazon reported for the period, and see any dollar differences with next actions. Read-only; fails closed on incomplete data.

amazon_reimbursement_audit

ChatGPT
Use this when the user asks whether Amazon reimbursed an FBA inventory loss, whether units remain unresolved, or whether Amazon may still owe reimbursement for one ASIN, FNSKU, or seller SKU over a date window. Reports what is resolved, what Amazon may still owe, and what evidence is missing. This audits evidence; it does not autonomously file a Seller Support claim. Absent data is never invented.

amazon_reimbursement_case_draft

ChatGPT
Build a paste-ready Amazon Seller Support case draft from a reimbursement audit, ready for the seller to review and file. DRAFT ONLY — never submits; the seller files manually.

amazon_reimbursement_evidence_ingest

ChatGPT
Trigger one evidence-ingest run for the connected seller: request the FBA ledger-detail and reimbursements reports for a date window, then poll, download, parse, and persist the observations the reimbursement audit reads. Runs are idempotent per window and advance on the per-minute scheduler; this call only enqueues. Operator-initiated; never auto-retries.

amazon_report_recovery_plan

ChatGPT
Create an owner/admin approval request to bind a pending report job to an exact Amazon reportId. Requires actual human Seller Central confirmation of all reportOptions; provider getReport cannot prove those options. Does not modify Amazon.

amazon_report_recovery_status

ChatGPT
Inspect one tenant-bound ambiguous report reservation. This is not Amazon report data; pending means the provider creation outcome is uncertain and must not be blindly retried.

amazon_returns_request

ChatGPT
Request the official Amazon FBA Returns report for the seller-authorized marketplace. Report creation is an external control action but does not mutate seller catalog, inventory, orders, or pricing.

amazon_returns_results

ChatGPT
Read a requested Amazon FBA Returns report. Returns bounded seller-operational rows from the official Amazon report; customer comments and direct contact/address fields are omitted from MCP output.

amazon_sales_traffic_history

ChatGPT
Read durable product and account-date Sales & Traffic observations already ingested for this tenant and seller connection, including complete schema-validated Amazon metric sections.

amazon_sales_traffic_insights

ChatGPT
Use this when the user asks why Amazon sales, traffic, conversion, or Buy Box performance changed; which products are strongest or weakest; or how mobile-app traffic compares with browser traffic. It abstracts away report IDs and reads available Sales & Traffic evidence. If the required report is not available, return the missing-evidence state and recommend the report-request tool instead of inventing a cause.

amazon_sales_traffic_request

ChatGPT
Request Amazon Sales & Traffic analytics for a verified marketplace and date range. Returns a reportId to poll; source is GET_SALES_AND_TRAFFIC_REPORT.

amazon_sales_traffic_results

ChatGPT
Read a requested Amazon Sales & Traffic report. Returns pending state until Amazon finishes; when done returns real per-product and per-date sales, units, sessions, page views, Buy Box %, and Unit Session % with evidence metadata.

amazon_search_catalog_performance_request

ChatGPT
Request tenant-bound Amazon Search Catalog Performance (SCP) funnel data for a complete week, month, or quarter, optionally narrowed to ASINs.

amazon_search_catalog_performance_results

ChatGPT
Poll and read schema-validated tenant-bound SCP ASIN funnel rows: impressions, clicks, cart adds, purchases, conversion, and search-attributed sales.

amazon_search_listings

ChatGPT
Search and page through this tenant-bound seller account listings by SKU/ASIN/GTIN/FNSKU, relationship, timestamps, issue severity, or buyability/discoverability status. Returns up to 20 items per page; Amazon caps traversal at 1,000 items.

amazon_search_orders

ChatGPT
Read one authoritative page from Amazon Orders API 2026-01-01 for a created or updated time window. Returns order/item economics, fulfillment, promotions, cancellations and package facts. Buyer, recipient, tax-registration and payment datasets are intentionally not requested on this general-purpose tool.

amazon_search_query_performance_request

ChatGPT
Request tenant-bound Amazon Search Query Performance (SQP) data for exact ASINs and one complete week, month, or quarter.

amazon_search_query_performance_results

ChatGPT
Poll and read schema-validated tenant-bound SQP rows: query volume, impressions, clicks, cart adds, purchases, and ASIN shares.

amazon_seller_performance_request

ChatGPT
Request the official Amazon Seller Performance report for the seller-authorized marketplace. Report creation is an external control action but does not mutate seller catalog, inventory, orders, or pricing.

amazon_seller_performance_results

ChatGPT
Read a requested Amazon Seller Performance report. Returns bounded seller-operational rows from the official Amazon report; customer comments and direct contact/address fields are omitted from MCP output.

amazon_seller_profile

ChatGPT
Read Amazon marketplace participation for the tenant-bound seller.

amazon_seller_pulse

ChatGPT
Compare the two most recent Amazon sales periods for the authenticated seller and marketplace. Returns account and SKU/ASIN sales, traffic, conversion, Featured Offer, device-traffic mix, trend series, ranked changes, evidence limits, and state-aware questions. It does not claim causation, device sales attribution, or landed profit.

amazon_storage_aging_report

ChatGPT
Use this when the user asks what their FBA inventory will cost under Amazon's unified age-based storage pricing (Feb 1 2027), which stock is about to cross the 181-day rate cliff, or what AWD stock aged over 12 months will cost. Buckets caller-supplied inventory lots by age band with the monthly cost per bucket under the published 2027 rates, and reports cubic feet crossing the 181-day cliff within 30/60/90 days. Published schedules only; ages and cubic feet are caller-supplied and Amazon's measurement governs billing.

amazon_storage_fee_request

ChatGPT
Request the official Amazon FBA Storage Fees report for the seller-authorized marketplace. Report creation is an external control action but does not mutate seller catalog, inventory, orders, or pricing.

amazon_storage_fee_results

ChatGPT
Read a requested Amazon FBA Storage Fees report. Returns bounded seller-operational rows from the official Amazon report; customer comments and direct contact/address fields are omitted from MCP output.

amazon_stranded_inventory_request

ChatGPT
Request the official Amazon Stranded Inventory report for the seller-authorized marketplace. Report creation is an external control action but does not mutate seller catalog, inventory, orders, or pricing.

amazon_stranded_inventory_results

ChatGPT
Read a requested Amazon Stranded Inventory report. Returns bounded seller-operational rows from the official Amazon report; customer comments and direct contact/address fields are omitted from MCP output.

amazon_support_navigator

ChatGPT
Classify seller support problems and return ranked source-graded, step-by-step evidence checks. Read-only; never creates a case, promises reimbursement, or changes Amazon state. Reference-only seller guidance: no support case created, no Amazon account read and no claimed reimbursement. Follow-up live tools require separate authorization.

amazon_support_playbook_lookup

ChatGPT
Read a curated evidence-graded problem playbook by ID. Sources are policy references, not proof of a seller outcome. Reference-only seller guidance: no support case created, no Amazon account read and no claimed reimbursement. Follow-up live tools require separate authorization.

amazon_support_topics

ChatGPT
List customer-facing Amazon Seller Support problem categories, without access to internal research, Seller Central cases, or seller financial data. Read-only reference topics.

ezseller_report_issue

ChatGPT
Create a private, tenant-bound and redacted EzSeller support report after the user asks to report a problem. Never include credentials, customer records, raw provider bodies, URLs with query strings, or private reasoning.

get_agency_billing_summary

ChatGPT
Read the tenant's agency billing summary: the plan and subscription status from the actual subscription record (the same facts as get_subscription_status), seller connections used against the plan's connection cap with the per-client split, and operator seats used against the plan's seat cap. Agency billing is a single subscription for the whole tenant; usage past the Floor limit is flagged as Agency Custom, which is scoped with the EzSeller team and has no checkout plan. A caller without an active billing membership gets a refusal, not data.

get_subscription_status

ChatGPT
Read the authenticated account's EzSeller subscription facts: billing provider (from the actual subscription record), plan, renewal on/off, paid-access end, whether further charges can occur, any pending change, and the actions this account is permitted to take. Works even when the paid entitlement or question quota has expired — billing management is not behind the paid-work gate.

seller_bi_ads_coverage_history

ChatGPT
Show months and rows covered by original Amazon advertising reports. NOT ad spend, sales attribution, ROAS, or live Ads API access.

seller_bi_compare_finance_months

ChatGPT
For Performix US, compare two exact settlement-posting months with source-pinned sales, signed Amazon fees, contribution and separate reimbursement postings. Returns dollars and changes. DOES NOT establish COGS, full Ads spend, shipping, bank payout, net profit or root causes.

seller_bi_finance_history

ChatGPT
Read source-verified, historical monthly Amazon settlement finance aggregates; not live finances or profit. USD minor units, no buyer records.

seller_bi_sku_history

ChatGPT
Read source-verified historical SKU sales/contribution aggregates with explicit missing COGS and advertising allocations. Not live inventory.

seller_bi_source_coverage_history

ChatGPT
Show Performix US historical source families, month coverage and explicitly missing evidence. No raw documents, case text, customer data, or Ads data.

submit_cancellation_feedback

ChatGPT
Record optional feedback about a cancellation. Feedback is never required to cancel, never delays a cancellation, and never changes the outcome of one. Safe to call before, during, or after a cancellation — or on its own.

Capabilities

Writes

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/ezseller"><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/ezseller)