extract-axure-data

Warn

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The script fetches JavaScript files (e.g., data/document.js and files/{page}/data.js) from user-provided URLs and executes them using the node:vm module's runInContext method.
  • Evidence: Found in scripts/extract.mjs within the evaluateAxureCode function where vm.runInContext(jsCode, sandbox) is called on code retrieved via fetchText.
  • [COMMAND_EXECUTION]: The skill uses execSync to automatically install software and dependencies on the host machine during its first run.
  • Evidence: In scripts/extract.mjs, the ensureDependencies function executes npm install playwright and npx playwright install chromium using execSync.
  • [EXTERNAL_DOWNLOADS]: The skill automatically downloads and installs the Playwright library and the Chromium browser from external registries and mirrors.
  • Evidence: scripts/extract.mjs references https://npmmirror.com/mirrors/playwright as a fallback mirror for downloading browser binaries.
  • [DYNAMIC_EXECUTION]: The script dynamically imports the Playwright library from a path constructed at runtime in the user's home directory.
  • Evidence: loadPlaywright in scripts/extract.mjs uses import() on a path built using os.homedir() and .cache/axure-extractor.
  • [COMMAND_EXECUTION]: The skill launches the Chromium browser with security features disabled, which can increase the risk of exploitation when navigating to untrusted sites.
  • Evidence: getBrowser in scripts/extract.mjs uses arguments --disable-web-security and --allow-file-access-from-files.
Audit Metadata
Risk Level
MEDIUM
Analyzed
May 20, 2026, 06:31 AM
Security Audit — agent-trust-hub — extract-axure-data