prep-that-slide
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from users (source notes, audience descriptions) and existing documents (attached PowerPoint files). This data is processed to generate presentation content, which creates a potential surface for indirect prompt injection if the inputs contain malicious instructions.
- Ingestion points: User prompts, attached
.pptxfiles (as described inreferences/slide-production.md), and content retrieved from Google Slides through authorized integrations. - Boundary markers: The instructions do not explicitly mandate the use of strict delimiters or boundary markers for all ingested text, although it advises distinguishing between observed facts and interpretations.
- Capability inventory: The skill utilizes file system operations (via agent tools to read/write PPTX, PDF, HTML), browser automation (for layout checks via
scripts/check-layout.js), and potentially the Google Slides API through authorized integrations. - Sanitization: No specific technical sanitization or escaping mechanisms for input strings are defined in the instructions before they are interpolated into the generated slide content.
Audit Metadata