frontend-slides
Pass
Audited by Gen Agent Trust Hub on Jul 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes platform-specific shell commands to launch the user's default browser to view the generated presentation, including
openfor macOS,xdg-openfor Linux, andstartfor Windows. - [EXTERNAL_DOWNLOADS]: The skill requires the
python-pptxpackage from the official Python Package Index (PyPI) to handle the parsing and conversion of PowerPoint files. - [REMOTE_CODE_EXECUTION]: The skill generates and executes local Python scripts using
python3to extract text, notes, and assets from.pptand.pptxfiles during the conversion process. - [PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it processes untrusted data from PowerPoint files and interpolates it into generated HTML/JS code.
- Ingestion points: External
.pptand.pptxfiles are ingested and parsed using thepython-pptxlibrary as defined inSKILL.md. - Boundary markers: There are no explicit delimiters or boundary markers defined in the instructions to prevent the agent from obeying instructions embedded within the extracted presentation content.
- Capability inventory: The skill possesses capabilities for local Python execution, file system writes (generating
.htmlfiles), and shell command execution to open files. - Sanitization: The instructions do not include requirements for sanitizing or escaping the text and metadata extracted from the source files before it is included in the final HTML output.
Audit Metadata