marketing-loops
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for loops that ingest data from external sources, such as social media mentions, news feeds (newsjacking), and customer feedback. While this creates a theoretical surface area for indirect prompt injection, the skill proactively mitigates this risk by mandating human review for all outbound or financial actions, implementing explicit veto lists for sensitive content, and providing detailed instructions on validating external signals before processing.
- [DATA_EXPOSURE]: The skill includes comprehensive guidance on handling sensitive data. It explicitly instructs against logging raw PII (Personally Identifiable Information) in state files or run logs, recommending the use of internal IDs or hashes and adhering to data minimization principles (GDPR/CCPA compliance).
- [COMMAND_EXECUTION]: The skill references platform-native scheduling primitives (such as /loop, ScheduleWakeup, and CronCreate) to manage automation cadences. These are documented as standard environment features rather than risky or unauthorized command execution patterns.
Audit Metadata