product-listing
Warn
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The instructions recommend running
uvx marketplaces-mcp-ruif the server is not present, which downloads a package from a remote registry to provide the necessary MCP server functionality. - [REMOTE_CODE_EXECUTION]: The use of
uvxto execute themarketplaces-mcp-rupackage facilitates the execution of remote code on the local system. This package is not from a verified trusted organization or explicitly linked to the author's namespace. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from external e-commerce platforms, which could contain adversarial instructions designed to influence the agent.
- Ingestion points: Data is retrieved from Wildberries, Ozon, and Yandex Market via tools like
wb_content_cards_list,ozon_product_list, andym_get_offer_mappings(SKILL.md). - Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are present when processing listing data.
- Capability inventory: The skill has the ability to write/update listings via
wb_content_cards_upload,wb_content_cards_update, andym_update_offer_mappings(SKILL.md). - Sanitization: There is no evidence of sanitization or validation of the data retrieved from external marketplaces before it is used to generate or update listings.
Audit Metadata