slide-deck-builder
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill utilizes well-known technology packages including 'python-pptx' for deck generation and 'markitdown' (a Microsoft-developed utility) for document content extraction.
- [COMMAND_EXECUTION]: Executes local shell scripts and Python scripts to automate the build and rendering process. Specifically, 'scripts/render_check.sh' uses 'soffice' (LibreOffice) and 'pdftoppm' for visual verification of generated files.
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves ingesting and processing untrusted data from external files (PDF, DOCX, transcripts, Notion pages) which presents an attack surface for indirect prompt injection.
- Ingestion points: Accepts various file formats and transcripts as primary content sources in the 'Ingest the source' step.
- Boundary markers: Instructions do not specify the use of delimiters or escaping when interpolating extracted source text into the agent's context.
- Capability inventory: The skill has the capability to write files (JSON outlines, PPTX decks) and execute shell commands for conversion and rendering.
- Sanitization: Relies on external extraction tools like 'markitdown' without additional validation or filtering of the extracted text for malicious instructions.
Audit Metadata