hai-visual-report

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The generated HTML reports load resources from external CDNs, specifically Tailwind CSS (cdn.tailwindcss.com) and Mermaid.js (cdn.jsdelivr.net). These are well-known technology services used for styling and rendering diagrams.
  • [COMMAND_EXECUTION]: The skill instructs the agent to create and save a .html file to the local file system, defaulting to the system temporary directory ($TMPDIR) or a path specified by the user.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. It processes untrusted data such as ideas, requirements, and reviews and incorporates it into the final HTML output.
  • Ingestion points: User-supplied descriptions of projects, requirements, or reviews as specified in SKILL.md.
  • Boundary markers: Absent; there are no instructions to use delimiters or warnings for the ingested content.
  • Capability inventory: Writing files to the local disk as described in SKILL.md.
  • Sanitization: Not present; the instructions do not mandate escaping or validating content before it is placed into the HTML skeleton template in references/html-skeleton.md.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 02:50 AM
Security Audit — agent-trust-hub — hai-visual-report