pptx-generator

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest brand guidelines, tone documents, and configuration files to customize slide generation. 1. Ingestion points: Reads .claude/skills/pptx-generator/brands/*/brand.json and associated .md files. 2. Boundary markers: The skill lacks explicit boundary markers or instructions to ignore embedded commands within these data files. 3. Capability inventory: Uses uv run to execute Python scripts and calls external tools like libreoffice or soffice. 4. Sanitization: There is no evidence of sanitization of brand values before they are interpolated into Python heredocs for execution.
  • [DYNAMIC_EXECUTION]: The core logic involves generating Python scripts at runtime based on brand data and executing them via uv run. This pattern is used for slide generation, batch processing, and merging part files into a final presentation.
  • [COMMAND_EXECUTION]: The skill invokes several shell commands for environment management and output generation, including mkdir and rm for directory operations, uv run for Python execution, and libreoffice or soffice for PDF conversion.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 03:42 PM