riskiest-assumption-test

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed to manage business logic within a project-specific directory. It reads a validation-canvas.md file and generates reports in Markdown and HTML formats.
  • [COMMAND_EXECUTION]: File system operations are restricted to reading and writing project artifacts. The skill explicitly uses safe practices such as atomic writes for manifest updates and preserving existing results in files.
  • [PROMPT_INJECTION]: While the skill processes user-provided content from the validation canvas, it mitigates risk by expecting a specific structure and including a built-in escapeHtml function in the generated HTML dashboard to prevent cross-site scripting (XSS).
  • [EXTERNAL_DOWNLOADS]: All external links point to official assets and installation paths on GitHub, which is a well-known and trusted service for development resources.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 07:12 AM
Security Audit — agent-trust-hub — riskiest-assumption-test