linkfox-sorftime-amazon-product-detail
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill communicates with several vendor-owned domains (linkfox.com) to retrieve Amazon product data, manage user authentication, and submit feedback. These include 'tool-gateway.linkfox.com', 'api.linkfox.com', 'agent-api.linkfox.com', and 'skill-api.linkfox.com'.
- [COMMAND_EXECUTION]: The skill executes local Python scripts ('scripts/sorftime_product_detail.py' and 'scripts/onboarding.py') to perform data retrieval and account management tasks.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external Amazon listings (titles, descriptions, bullet points) through the Sorftime API.
- Ingestion points: API response 'products' array in 'scripts/sorftime_product_detail.py'.
- Boundary markers: The instructions advise the agent to present data in clear tables and summarize large responses, but no specific 'ignore instructions' delimiters are enforced on the raw text fields.
- Capability inventory: The skill has the ability to write JSON data files to the local file system ('linkfox/' directory).
- Sanitization: The scripts parse the structured JSON response, but text fields within the JSON are not sanitized for potential natural language instructions before being presented to the agent.
Audit Metadata