Skills
skills/yysun/apprun/apprun-skills/Gen Agent Trust Hub

apprun-skills

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The playground component in src/apprun-code.tsx and src/apprun-play.tsx dynamically executes user-provided code snippets using document.write and ts.transpileModule within an iframe sandbox.
  • [REMOTE_CODE_EXECUTION]: The calculator demonstration (docs/_src/calculator.js and demo/components/calculator.tsx) utilizes eval() to process arithmetic expressions constructed from component state.
  • [EXTERNAL_DOWNLOADS]: The skill fetches demonstration data and library dependencies from well-known and trusted services, including Netlify (xkcd-api.netlify.app), Glitch (my-xkcd-api.glitch.me), and CDNs like unpkg and jsDelivr.
  • [PROMPT_INJECTION]: An indirect prompt injection surface exists in the playground component (src/apprun-code.tsx). Untrusted code data enters the context via DOM element extraction without strict boundary markers or sanitization, enabling dynamic execution through the framework's internal rendering capabilities. This risk is inherent to the component's purpose as a code execution tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 10:01 PM
Security Audit — agent-trust-hub — apprun-skills