linkfox-mpstats-ozon-product-trend
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts (
mpstats_ozon_product_trend.pyandonboarding.py) to retrieve trend data and handle authentication flows. This behavior is necessary for the skill's documented functionality. - [DATA_EXFILTRATION]: The skill transmits parameters and authentication identifiers to several subdomains of
linkfox.com(such astool-gateway.linkfox.comandapi.linkfox.com). This communication is directed to the vendor's own infrastructure for legitimate API operations. - [INDIRECT_PROMPT_INJECTION]: The skill ingests marketplace data from external APIs, which serves as a potential attack surface. However, the data consists of strictly formatted numeric values and dates (e.g., prices, sales counts, inventory levels), which minimizes the risk of executing embedded instructions.
- Ingestion points: Trend data fetched from the LinkFox gateway in
mpstats_ozon_product_trend.py. - Boundary markers: No explicit delimiters are used when the data is interpolated into the agent's context.
- Capability inventory: The skill can write files to the local
linkfox/project directory and perform network POST requests to vendor APIs. - Sanitization: API responses are parsed as JSON and processed through specific numeric and string fields.
Audit Metadata