google-ads-keywords-for-keywords
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The execution flow is well-structured and secure. The bash script uses quoted here-docs (<<'PY') to execute inline Python code, which prevents shell variable expansion vulnerabilities. Temporary files are managed with mktemp and cleaned up on exit using a trap. The network connection to the vendor API (ai-factory.frevana.com) is for core functionality and does not involve sensitive data exfiltration. The skill allows saving output to a user-specified path, which is a standard feature for this type of tool.
- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes external API data.
- Ingestion points: The response from the ai-factory.frevana.com API is processed in scripts/search_google_ads_keywords_for_keywords.sh.
- Boundary markers: None identified; the instructions suggest returning the data unchanged.
- Capability inventory: Includes network access (curl), file system writes (mkdir, cp), and Python execution.
- Sanitization: The script validates that the response is JSON but does not perform sanitization on the data values within the JSON object.
Audit Metadata