google-ads-account-audit
Pass
Audited by Gen Agent Trust Hub on Mar 24, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No instructions were found that attempt to bypass AI safety guidelines, disregard previous rules, or simulate restricted modes. The skill remains within its defined scope of a Google Ads auditor.
- [DATA_EXFILTRATION]: While the skill suggests gathering account context and data, it contains no mechanisms (such as network requests or file writes) to exfiltrate this information. It is designed for manual data analysis within the agent's session.
- [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or credentials are present. The skill instructs users on where to find information within the Google Ads UI rather than requesting programmatic access via secrets.
- [EXTERNAL_DOWNLOADS]: The skill does not attempt to download external scripts, tools, or configuration files.
- [REMOTE_CODE_EXECUTION]: There is no evidence of code execution patterns, such as piping remote content to a shell or using dynamic evaluation of untrusted strings.
- [COMMAND_EXECUTION]: The skill does not invoke shell commands, system utilities, or subprocesses.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for indirect injection by ingesting Google Ads exports or screenshots. However, because the agent lacks execution capabilities (e.g., no tool access defined), the risk is negligible. 1. Ingestion points: User-provided ads data/exports. 2. Boundary markers: None. 3. Capability inventory: None (text output only). 4. Sanitization: None.
Audit Metadata