yandex-market-mcp-ru
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| YANDEX_MARKET_API_KEY | No | API-ключ из кабинета партнёра, раздел Настройки → API. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| ym_check_authA | Check whether the required credentials are present in the environment. Does NOT reveal secret values — only reports which variables are set. Returns JSON: {"ready": bool, "missing": [str], "required": [str]}. |
| ym_list_sectionsA | List API sections and how many catalog endpoints each contains. |
| ym_get_sectionA | List all endpoints in one section. Args: section: section name (see ym_list_sections), e.g. "statistics". Returns JSON list of {operation_id, method, path, safety, summary}. |
| ym_search_methodsA | Search the endpoint catalog by keyword (works in Russian and English). Args: query: free text, e.g. "остатки", "stocks", "update price". limit: max results (1-50). Returns JSON list of matching endpoints (best first). |
| ym_mapA | The big picture: business entities this API covers and the go-to methods for each. Call with no args to see the whole map ("you are here"); pass entity="reviews" (or stocks/prices/orders/…) to list every method of one entity. Use this before guessing — it orients you fast. |
| ym_describe_methodA | Return the full catalog record for one endpoint: method, host, path, scope, safety level, pagination style, rate limit, params and doc URL. |
| ym_call_methodA | Execute one READ endpoint from the catalog by operation_id. Target API: https://yandex.ru/dev/market/partner-api/doc/ru/. Reads only: nothing here changes data, so it runs without confirmation. To change data use ym_write_method, to delete use ym_delete_method. Args: operation_id: id from the catalog (see ym_search_methods). path_values: values for {placeholders} in the path. query: query-string parameters. body: JSON request body (a few read endpoints take one). Returns JSON: {"ok": true, "status", "data"} or the error envelope. |
| ym_write_methodA | Execute one WRITE endpoint from the catalog: create or update data. Target API: https://yandex.ru/dev/market/partner-api/doc/ru/. Requires confirm_write=true; nothing is sent without it. Irreversible operations live in ym_delete_method, reads in ym_call_method. Args: operation_id: id from the catalog (see ym_search_methods). path_values: values for {placeholders} in the path. query: query-string parameters. body: JSON request body. confirm_write: must be true. Returns JSON: {"ok": true, "status", "data"} or the error envelope. |
| ym_delete_methodA | Execute one DESTRUCTIVE endpoint: deletes or irreversibly changes data. Target API: https://yandex.ru/dev/market/partner-api/doc/ru/. Both confirm_write=true and i_understand_this_modifies_data=true are required; nothing is sent without both. Args: operation_id: id from the catalog (see ym_search_methods). path_values: values for {placeholders} in the path. query: query-string parameters. body: JSON request body. confirm_write: must be true. i_understand_this_modifies_data: must be true. Returns JSON: {"ok": true, "status", "data"} or the error envelope. |
| ym_get_rawA | Read ANY endpoint by path, including ones missing from the catalog. Target API: https://yandex.ru/dev/market/partner-api/doc/ru/. Safe verbs only (GET, HEAD, OPTIONS). To change data use ym_write_raw, to delete use ym_delete_raw. Args: path: full path beginning with '/', e.g. "/v2/regions". method: safe verb, GET by default. host: host override; defaults to the service's default host. query: query-string parameters. body: JSON request body (rare on reads; some APIs want one). Returns JSON: {"ok": true, "status", "data"} or the error envelope. |
| ym_write_rawA | Create or update data at ANY path, including paths not in the catalog. Target API: https://yandex.ru/dev/market/partner-api/doc/ru/. POST, PUT and PATCH only; requires confirm_write=true. Args: method: POST, PUT or PATCH. path: full path beginning with '/'. host: host override; defaults to the service's default host. query: query-string parameters. body: JSON request body. confirm_write: must be true. Returns JSON: {"ok": true, "status", "data"} or the error envelope. |
| ym_delete_rawA | Delete data at ANY path, including paths not in the catalog. Target API: https://yandex.ru/dev/market/partner-api/doc/ru/. DELETE only. Both confirm_write=true and i_understand_this_modifies_data=true are required. Args: path: full path beginning with '/'. method: DELETE. host: host override; defaults to the service's default host. query: query-string parameters. body: JSON request body. confirm_write: must be true. i_understand_this_modifies_data: must be true. Returns JSON: {"ok": true, "status", "data"} or the error envelope. |
| ym_fetch_allA | Auto-paginate a read endpoint and return every row in one response. Handles offset, last_id, cursor (Ozon v4/v5), page and WB lastChangeDate styles. The array path is taken from the catalog automatically. Args: operation_id: a read endpoint from the catalog. query / body / path_values: base parameters (cursor fields are managed). items_path: override the array path (default: the endpoint's own). limit: page size to request. max_items: hard cap to protect context (default 10000). Returns JSON: {"ok", "items", "total_fetched", "pages_fetched", "truncated"}. |
| ym_list_cabinetsA | List configured cabinets for this marketplace and which one is active. Returns JSON: {"active": str|null, "cabinets": [names], "fields_needed": [...]}. Secret values are never returned. |
| ym_add_cabinetA | Add or update a cabinet (a named set of API credentials), from chat. ⚠️ This puts the key into the chat transcript — requires i_understand_key_goes_to_chat=true. The terminal-free safe alternative is the installer (install.py / double-click), where the key never enters chat. Args: credentials: dict with the required fields for this service ({fields}). For Ozon: {{"client_id": "...", "api_key": "..."}}; for WB: {{"token": "..."}}. name: optional label. If omitted, the cabinet is named after the real shop name fetched from the marketplace (falls back to "main"). i_understand_key_goes_to_chat: must be true to proceed. Saved to ~/.marketplace-mcp/cabinets.json (local, chmod 600), never echoed. |
| ym_set_keyA | Change / rotate the API key from chat (e.g. the old one expired or leaked). ⚠️ The key goes into the chat transcript — requires i_understand_key_goes_to_chat=true. The safe, terminal-free alternative is the installer, where the key never enters chat. Use a scoped key and rotate it in the seller cabinet if it was exposed. Args: credentials: dict with the required fields ({fields}). cabinet: which cabinet to update. Default: the active one (so "my key expired" just works). If there is none, the cabinet is named from the marketplace's shop name, else "main". i_understand_key_goes_to_chat: must be true to proceed. On success the key is validated against the marketplace and the shop name is reported. Saved locally (chmod 600), never echoed back. |
| ym_use_cabinetA | Switch the active cabinet. Subsequent API calls use its credentials. Args: name: the cabinet to activate (see ym_list_cabinets). |
| ym_remove_cabinetA | Delete a stored cabinet. If it was active, another becomes active. Args: name: the cabinet to remove. |
| ym_list_workflowsA | List ready-made analytical workflows (recipes) for this marketplace. Returns JSON: [{name, category, when_to_use}]. Use {svc}_get_workflow to fetch the full step-by-step plan for one. |
| ym_get_workflowA | Return the full plan for one workflow: ordered steps (each naming a catalog operation_id and why), interpretation guidance, and common mistakes to avoid. Args: name: workflow name (see {svc}_list_workflows). |
| ym_get_campaignsA | List the seller's shops (кампании) with their campaignId and businessId. Call this first: nearly every other Yandex Market method needs a campaignId (магазин) or businessId (кабинет продавца). Both are in the response: campaigns[].id and campaigns[].business.id. Args: page_token: pageToken from a previous page (empty for the first page). limit: page size (<=100). Returns JSON: {"ok": true, "data": {"campaigns": [...], "paging": {...}}}. |
| ym_get_ordersA | List orders of one shop (GET /v2/campaigns/{campaignId}/orders). Args: campaign_id: shop id from ym_get_campaigns. status: filter, e.g. PROCESSING | DELIVERY | PICKUP | DELIVERED | CANCELLED | UNPAID (comma-separated allowed). Empty = all. from_date: order creation date lower bound, DD-MM-YYYY (Yandex format). to_date: upper bound, DD-MM-YYYY. page_token: pageToken from a previous page. limit: page size (<=50). Returns JSON: {"ok": true, "data": {"orders": [...], "paging": {...}}}. For every order across pages use ym_fetch_all with ym_get_orders. |
| ym_get_offersA | List the seller's offers (товары) with their Market card mapping (POST /v2/businesses/{businessId}/offer-mappings). Args: business_id: cabinet id (campaigns[].business.id). offer_ids: comma-separated offerId (SKU) filter; empty = all. page_token: pageToken from a previous page. limit: page size (<=200). Returns JSON: {"ok": true, "data": {"result": {"offerMappings": [...], "paging": {...}}}}. |
| ym_get_stocksA | Stock per offer per warehouse for one shop, with optional turnover (POST /v2/campaigns/{campaignId}/offers/stocks). Args: campaign_id: shop id. offer_ids: comma-separated offerId filter; empty = all. with_turnover: also return turnover (оборачиваемость) per offer. page_token: pageToken from a previous page. limit: page size (<=200). Returns JSON: {"ok": true, "data": {"result": {"warehouses": [{"warehouseId", "offers": [...]}]}}}. |
| ym_get_pricesA | Base prices set for all shops of the cabinet (POST /v2/businesses/{businessId}/offer-prices). Args: business_id: cabinet id. offer_ids: comma-separated offerId filter; empty = all. page_token: pageToken from a previous page. limit: page size (<=200). Returns JSON with result.offers[].price {value, currencyId, discountBase, updatedAt}. |
| ym_set_priceA | Set the base price of ONE offer for all shops of the cabinet (POST /v2/businesses/{businessId}/offer-prices/updates). WRITE. Requires confirm_write=true. discount_base is the strikethrough price (must be higher than price); 0 = no discount shown. Args: business_id: cabinet id. offer_id: seller's SKU (offerId). price: new price, e.g. 1499. discount_base: pre-discount price, or 0 to clear. currency: RUR (default) — Yandex uses "RUR", not "RUB". confirm_write: must be true to send. Returns JSON: {"ok": true, "data": {"status": "OK"}} on success. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 26 tools
The tools fall into clearly delineated groups: catalog discovery, generic catalog execution, raw-path execution, cabinet management, workflows, and domain convenience methods. Even paired tools like ym_call_method/ym_get_raw and ym_write_method/ym_write_raw are cleanly separated by catalog-based versus raw-path targeting, so an agent should not misselect.
All tools share the ym_ prefix and follow a consistent verb_noun or verb_adjective pattern, such as list_sections, get_offers, set_price, remove_cabinet, call_method, and write_raw. The few one-word names (ym_map, ym_fetch_all) still fit the overall style, and there is no mixing of camelCase or vague verbs.
At 26 tools, the server is above the ideal range and slightly heavy. However, the count is justified by the broad scope: catalog exploration, generic and raw API execution, cabinet management, workflows, and convenience endpoints all serve distinct purposes. It feels like a large but deliberate surface rather than bloated.
The generic catalog-based and raw-path executors cover read, write, and delete operations, meaning the entire Yandex Market Partner API is reachable and there are no dead ends. Convenience methods cover the core seller workflows (campaigns, offers, orders, stocks, prices), and cabinet/auth setup is fully handled, so the surface is effectively complete.