linkfox-shopee-store-returns
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes
subprocess.runwithinscripts/_shopee_returns_common.pyto execute a local helper script,check_auth_dependency.py. This is performed to verify that required prerequisite skills are present in the environment. - [INDIRECT_PROMPT_INJECTION]: The skill manages a surface for indirect prompt injection by processing external data from Shopee APIs and handling user-supplied JSON payloads.
- Ingestion points: User-provided JSON parameters passed to entry scripts like
returns_api.pyand individual API wrappers. - Boundary markers: The instructions do not define explicit delimiters or warnings to the model regarding potential instructions embedded in the API data.
- Capability inventory: The skill has the capability to perform network requests to vendor APIs and execute local validation scripts via
subprocess. - Sanitization: Input parameters are parsed as JSON objects, which provides basic structural validation but does not filter for malicious natural language instructions.
Audit Metadata