frontend-slides
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the
python-pptxandPillowpackages viapipto support PowerPoint conversion and image processing. These are well-known libraries from official package registries. - [COMMAND_EXECUTION]: The agent is instructed to execute a local Python script,
scripts/extract-pptx.py, which is included as part of the skill's package. This script uses thepython-pptxlibrary to read and extract content from user-provided presentation files. - [DYNAMIC_EXECUTION]: The skill involves the generation of executable content at runtime:
- The core functionality generates a self-contained HTML file containing embedded CSS and JavaScript for slide navigation and interactive elements.
- The
html-template.mdfile provides Python snippets intended for runtime execution to perform image manipulations like circular cropping and resizing. - [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it processes data from untrusted external files.
- Ingestion points: Content is ingested from user-provided PowerPoint files (.pptx) via the
extract-pptx.pyscript and from local image directories. - Boundary markers: There are no specific delimiters or instructions to ignore potential commands embedded within the extracted PPTX text.
- Capability inventory: The skill can execute local Python scripts and write new HTML files to the filesystem.
- Sanitization: The instructions do not explicitly mandate the sanitization or escaping of extracted text content before it is interpolated into the generated HTML slides.
Audit Metadata