ppt-keynote

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill references styling and typography assets from well-known and trusted services, specifically the Tailwind CSS CDN and Google Fonts.
  • [SAFE]: Client-side JavaScript within the generated slide template is restricted to benign keyboard event listeners for navigating between sections.
  • [PROMPT_INJECTION]: The skill functions as a template that ingests user-supplied data to populate HTML slides, which presents an indirect prompt injection surface.
  • Ingestion points: User-provided document content and metadata defined in the template instructions in SKILL.md.
  • Boundary markers: No explicit delimiters or instructions (e.g., 'ignore embedded commands') are provided to the agent to isolate user content from the slide generation logic.
  • Capability inventory: The skill configuration in open-design.json specifies 'fs:write' and 'prompt:inject' capabilities, allowing it to create files on the system and influence the agent's internal prompting logic.
  • Sanitization: The skill does not define specific validation, escaping, or filtering mechanisms for the external content it processes into the final HTML output.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 05:04 PM
Security Audit — agent-trust-hub — ppt-keynote