short-form-video-script
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external sources including delivery briefs, brand profiles, and voice builders (identified in SKILL.md), which serves as an attack surface for indirect prompt injection.
- Ingestion points: Untrusted data enters the agent context via 'delivery brief', 'brand-profile', and 'voice-builder' inputs.
- Boundary markers: The instructions lack explicit requirements for delimiters or 'ignore embedded instructions' warnings when handling these external inputs.
- Capability inventory: The skill coordinates with powerful downstream capabilities such as media generation ('ai-video') and social media publishing ('WoopSocial'), as documented in references/scope-and-connections.md.
- Sanitization: There is no mention of validation, escaping, or filtering of the content retrieved from external briefs and profiles.
- [COMMAND_EXECUTION]: The skill functions as an orchestrator that chains user inputs and external data to drive actions in various other skills and services (e.g., ai-video, reels-script, WoopSocial). This capability chain increases the potential impact if the scripting process is compromised by malicious input.
Audit Metadata