tt-caption-writer
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input and external configuration files, creating a surface for indirect prompt injection.\n
- Ingestion points: The skill loads
../../references/voice-profile.mdand prompts the user for video topics and scripts in the 'Gather inputs' step.\n - Boundary markers: Absent. There are no instructions to the agent to treat the content of the voice profile or user script as data rather than instructions.\n
- Capability inventory: The skill utilizes
lib/publora_client.pyto calllib.publish(), which performs network operations to external services (Publora/TikTok), andlib.illustrate()which calls an external image generation service.\n - Sanitization: Absent. While a 'Humanizer pass' is mentioned in
SKILL.mdto remove AI-like vocabulary patterns, no security-related sanitization or escaping of the user-provided script is performed before it is passed to the publishing function.
Audit Metadata