yandex-market-mcp

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install the yandex-market-mcp-ru package directly from the author's GitHub repository (github.com/ilyautov/yandex-market-mcp-ru) using the uvx tool.
  • [REMOTE_CODE_EXECUTION]: The installation process involves executing code retrieved from a remote Git repository to set up the MCP server environment.
  • [INDIRECT_PROMPT_INJECTION]: The skill interacts with the Yandex Market API, processing external data such as catalog information and order details.
  • Ingestion points: Data enters the context via the ym_call_method tool which fetches live data from Yandex Market.
  • Boundary markers: The skill includes explicit instructions for the agent to show the request body to the user and wait for confirmation before performing any write operations.
  • Capability inventory: The skill can search, describe, and call API methods, which includes both read and write capabilities (e.g., changing prices or managing orders).
  • Sanitization: The instructions advise the agent not to round or recalculate numbers and to report empty lists accurately, ensuring data integrity.
  • [CREDENTIALS_UNSAFE]: The skill requires a YANDEX_MARKET_API_KEY. It instructs users to provide this via environment variables or a local configuration file (~/.marketplace-mcp/cabinets.json). It uses a placeholder <значение> and does not contain hardcoded secrets.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 02:16 PM
Security Audit — agent-trust-hub — yandex-market-mcp