skills/simota/agent-skills/ink/Gen Agent Trust Hub

ink

Pass

Audited by Gen Agent Trust Hub on Oct 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest user-supplied requirements and existing icon style guides to generate SVG code and frontend components. This creates a potential surface for indirect prompt injection if the source data contains adversarial instructions. Ingestion points include user requirements for icons and illustrations, and style direction for icon systems as described in SKILL.md. The instructions emphasize structural optimization and accessibility but do not explicitly define delimiters for untrusted data or provide instructions to ignore embedded commands in source SVGs. The skill has the capability to write files (.svg, .tsx, .vue) and provides documentation for executing CLI tools (svgo, svgstore). While the skill utilizes SVGO for path optimization and metadata removal, this is focused on file size and performance rather than security sanitization of instructional content.
  • [SAFE]: The skill references established standards and documentation from trusted organizations such as the W3C (w3.org), Google (developers.google.com), and the International Organization for Standardization (iso.org). These references are used for legitimate technical guidance.
  • [SAFE]: The skill documentation includes examples of using industry-standard development tools like svgo and svgstore via the npx package runner. These are standard practices for SVG asset optimization and sprite generation in web development environments.
  • [SAFE]: All external URLs and resources point to well-known services and libraries such as Lucide, Heroicons, FontAwesome, and Smashing Magazine, which are widely recognized and trusted in the developer community.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 22, 2026, 01:03 PM