linkfox-keepa-product-search
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEPERSISTENCEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [PERSISTENCE]: The onboarding guide in
references/onboarding.mdprovides instructions for the user to add environment variable export commands to shell profile files such as~/.bashrcand~/.zshrcto ensure theLINKFOX_AGENT_API_KEYpersists across sessions.\n- [EXTERNAL_DOWNLOADS]: The skill's scripts (keepa_product_search.pyandonboarding.py) perform network operations targeting several subdomains belonging to the vendor (tool-gateway.linkfox.com,api.linkfox.com,agent-api.linkfox.com, andskill-api.linkfox.com) to facilitate product searches, authentication, and feedback reporting.\n- [INDIRECT_PROMPT_INJECTION]:\n - Ingestion points: The skill retrieves external data such as product titles, brands, and descriptions from the Keepa database through the search API.\n
- Boundary markers: There are no specific instructions or delimiters provided to ensure the agent ignores potential malicious instructions embedded within the retrieved product descriptions.\n
- Capability inventory: The skill executes local Python scripts capable of performing network requests and writing to the local file system within the workspace.\n
- Sanitization: The fetched product data is processed and presented to the agent without evidence of sanitization or filtering to prevent indirect injection attacks.
Audit Metadata