ppt-template-creator
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a process for ingesting external PowerPoint files (.pptx/.potx) provided by users to extract structural layout data. This constitutes a potential attack surface for indirect prompt injection, although the current implementation focuses on extracting specific structural attributes.
- Ingestion points: Processes user-provided template files using the
python-pptxlibrary within the Python code blocks provided inSKILL.md. - Boundary markers: The skill does not explicitly use delimiters or markers to separate the binary file content from the agent context, relying on the library's parsing logic.
- Capability inventory: The agent is instructed to perform file system operations (reading templates) and code generation (writing new skill folders and files).
- Sanitization: The analysis script specifically targets numeric coordinates (left, top, width, height) and enum-based placeholder types, which effectively filters out arbitrary content or embedded instructions for the purpose of structural mapping.
- [DYNAMIC_EXECUTION]: The skill's primary function is to generate new Python code and markdown files. It uses static templates to create instructions for a new skill's
SKILL.md. This constitutes script generation from known templates, which is a legitimate automation task and does not involve the dangerous evaluation of untrusted external code strings in the current execution environment.
Audit Metadata