paid-media
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines workflows that ingest untrusted data from external advertising platforms and use that data to perform automated account modifications.\n
- Ingestion points: Reads search query reports, competitor ad data, and campaign performance metrics via external APIs and MCP tools as described in the specialist references (e.g., search-query-analyst.md and auditor.md).\n
- Boundary markers: Absent. The instructions do not define delimiters or specific safety markers to prevent the agent from obeying instructions embedded within external data strings.\n
- Capability inventory: The skill describes capabilities for account-level write operations, including creating campaigns, modifying bid strategies, and deploying negative keywords (referenced in ppc-strategist.md and search-query-analyst.md).\n
- Sanitization: Not present. There are no requirements to sanitize or validate strings from search queries or creative assets before they are used to drive account updates.\n- [NO_CODE]: The skill package consists entirely of markdown documentation files and role definitions. It does not include any scripts, binaries, or automated configuration files for execution on the host system.
Audit Metadata