google-ads-intent-map
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted search term data from Google Ads, creating a potential surface for indirect prompt injection where malicious instructions could be embedded in the data being processed.
- Ingestion points: The skill retrieves data via the
google-ads-mcptool, specifically queryingsearch_term_view.search_termas described inSKILL.md. - Boundary markers: The instructions do not define specific delimiters or "ignore embedded instructions" warnings for the agent when processing the results of the GAQL queries.
- Capability inventory: The skill is restricted to file system operations within the
workspace/directory (e.g., writing toworkspace/ads/intent-map.md). It does not possess network exfiltration tools, arbitrary shell execution, or privilege escalation capabilities. - Sanitization: No sanitization, validation, or filtering of the external search term content is specified before the data is processed by the agent.
Audit Metadata