The Agent Skills Directory

[SAFE]: The skill communicates with official vendor API endpoints (linkfox.com) to fetch brand statistics, which is consistent with its stated purpose.
[SAFE]: Data persistence is managed by writing results to a session-specific 'linkfox' directory in the workspace. This is used for caching and session history rather than unauthorized data storage.
[SAFE]: Authentication is managed via user-provided environment variables, following standard security practices for agent tools. The onboarding script facilitates this process by guiding the user through a manual setup.
[INDIRECT_PROMPT_INJECTION]: The skill ingests marketplace data from Ozon via the MPSTATS API, which presents a surface for indirect prompt injection.
Ingestion points: Product titles, category paths, and brand names are fetched from the API and processed in scripts/mpstats_ozon_brand_products.py.
Boundary markers: None identified; external data is not delimited with specific safety markers.
Capability inventory: The skill has the capability to execute Python scripts and write JSON files locally.
Sanitization: The skill summarizes and parses the data but lacks explicit sanitization of product metadata strings before they are returned to the agent context.

linkfox-mpstats-ozon-brand-products