skills/ilteoood/harness/slidev/Gen Agent Trust Hub

slidev

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill documents the use of the slidev CLI and pnpm for project lifecycle management, including initialization (pnpm create slidev), local development (pnpm run dev), production builds (pnpm run build), and slide exporting (pnpm run export).- [INDIRECT_PROMPT_INJECTION]: The skill provides structured tools via a built-in MCP server (slidev-update-slide, slidev-insert-slide, etc.) to read and modify Markdown files. This creates a surface where an agent might process untrusted instructions embedded in slide content when editing existing decks.
  • Ingestion points: Slide Markdown files (.md).
  • Capability inventory: Reading, updating, inserting, moving, and removing slides, as well as navigating a live presentation.
  • Boundary markers: Not explicitly defined in the skill documentation.
  • Sanitization: Relies on the underlying Slidev framework and the agent's own safety guardrails.- [EXTERNAL_DOWNLOADS]: The documentation references several external dependencies and services required for full functionality, such as playwright-chromium for PDF exports, various Iconify icon collections for styling, and remote diagram rendering services (PlantUML). These are standard components of the Slidev ecosystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 04:15 AM
Security Audit — agent-trust-hub — slidev