marp-slides
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill generates presentation content based on user-provided prompts. The generated Markdown is processed with the
--htmlflag enabled, which allows raw HTML and scripts to be rendered in the output. - Ingestion points: User prompts for slide deck requirements and content (SKILL.md).
- Boundary markers: No explicit boundary markers or instructions are provided to the agent to ignore potentially malicious instructions embedded in user-supplied content.
- Capability inventory: The skill invokes shell commands via
npxto executemarp-cli(SKILL.md, EXPORT.md), which renders the provided Markdown and HTML into slide formats. - Sanitization: No explicit sanitization or validation of the user-provided text is performed before it is written to the slide files.
- [COMMAND_EXECUTION]: The skill uses the system shell to execute the
marp-clitool vianpx. This is the primary functionality of the skill, used to convert Markdown source files into PDF, PPTX, or HTML presentations. - [EXTERNAL_DOWNLOADS]: The skill uses
npxto download and run the@marp-team/marp-clipackage from the NPM registry. It also references external resources from well-known services, including Google Fonts and Unsplash for themes and examples (THEMES.md, editorial.md).
Audit Metadata