campaign-delivery
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run a local script
scripts/promo_delivery.pyto reconcile campaign media and verify deliverables. - [INDIRECT_PROMPT_INJECTION]: The skill processes social media copy and deliverables which may come from untrusted external sources.
- Ingestion points: Social campaign copy, deliverables, and specifications (SKILL.md).
- Boundary markers: Absent; no specific delimiters or instructions to ignore embedded commands are present.
- Capability inventory: Invokes local script execution via
scripts/promo_delivery.py. - Sanitization: Absent; no sanitization or validation of the processed text is described.
Audit Metadata