linkfox-mpstats-ozon-product-detail
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes Python scripts (
scripts/mpstats_ozon_product_detail.pyandscripts/onboarding.py) to fetch marketplace metrics and manage authentication workflows. - [EXTERNAL_DOWNLOADS]: The scripts communicate with vendor-controlled domains (
tool-gateway.linkfox.com,api.linkfox.com,agent-api.linkfox.com) to retrieve product data and handle user onboarding. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted marketplace data (e.g., product titles and descriptions), creating a potential injection surface.
- Ingestion points: Data is retrieved from Ozon via API calls in
scripts/mpstats_ozon_product_detail.py. - Boundary markers: The instructions do not define specific delimiters or warnings to separate marketplace content from system instructions.
- Capability inventory: The skill possesses script execution capabilities and network access to vendor APIs.
- Sanitization: No explicit sanitization or filtering of external marketplace content is described before the data is presented to the agent.
- [SAFE]: The onboarding documentation (
references/onboarding.md) provides legitimate instructions for users to manually configure their environment by adding API keys to shell profiles (.zshrc,.bashrc).
Audit Metadata