linkfox-shopee-store-fbs
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
subprocess.runinscripts/_shopee_fbs_common.pyto execute a local utility script,check_auth_dependency.py, which validates the presence of required peer skills. The executed script path is programmatically determined relative to the calling file.\n- [INDIRECT_PROMPT_INJECTION]: The skill acts as a proxy for Shopee API data, creating a surface where untrusted external content could influence agent behavior.\n - Ingestion points: API responses are fetched via
developer_proxy_callinscripts/_fbs_api_runner.pyand processed inscripts/_shopee_fbs_common.py.\n - Boundary markers: Output is formatted as JSON, but the skill lacks specific delimiters or warnings to the agent to ignore instructions embedded within the API responses.\n
- Capability inventory: The skill possesses capabilities for network communication with vendor-controlled domains (*.linkfox.com), local file system operations (saving logs to a linkfox/ directory), and execution of local peer scripts.\n
- Sanitization: External data is serialized using
json.dumpsand large responses are summarized before being presented to the agent, which partially mitigates the risk of large-scale injection.
Audit Metadata