skills/nexscope-ai/nexscope-ecommerce-skills/ecommerce-amazon-ads-sp-insights-report/Gen Agent Trust Hub
ecommerce-amazon-ads-sp-insights-report
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The entry scripts
scripts/get_sp_audience_report.pyandscripts/get_sp_search_impression_share.pyutilizesubprocess.runto execute a local dependency checker scriptscripts/check_auth_dependency.py. This execution is used to verify the presence of a required authorization skill before proceeding with API operations.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external CSV reports fetched from the Amazon Ads API. Inscripts/reporting_v1_workflow.py, the_preview_csvfunction reads these files usingcsv.DictReaderand returns the content as a preview to the agent. This represents an indirect prompt injection surface if the external data contains malicious instructions.\n - Ingestion points:
_preview_csvinscripts/reporting_v1_workflow.pyreads data from files saved innexscope/.\n - Boundary markers: The skill does not employ specific delimiters to isolate the CSV data from the agent's instructions.\n
- Capability inventory: The skill performs authenticated network requests to
api.nexscope.aiand writes results to the local filesystem.\n - Sanitization: No specific content sanitization is performed on the CSV values beyond standard CSV parsing.\n- [DATA_EXFILTRATION]: The skill communicates with
api.nexscope.ai(Vendor Resource) to proxy Amazon Ads requests. It includes logic inscripts/reporting_v1_workflow.py(_redact_presigned_urls) to remove sensitive download parameters from logs and responses. Additionally,_reject_sensitive_valuesprevents the inclusion of Amazon Advertising API secrets or tokens in the skill's request parameters.
Audit Metadata