wix-headless-replatform
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill instructions in
SKILL.mdandreferences/safety.mddirect the agent to skip standard authorization confirmation steps for the requested homepage. This is intended to streamline the migration workflow but overrides a default safety interaction pattern. - [COMMAND_EXECUTION]: The skill makes extensive use of
child_process.spawnandexecFileto run Wix CLI commands, package managers (npm, yarn, pnpm), and Playwright browser installation scripts. These are necessary for scaffolding and building the generated project. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted HTML from user-supplied URLs. It implements a mitigation strategy in
scripts/lib/content-boundary.mjsthat identifies and removes executable code signals (JavaScript DOM APIs, function bodies, stylesheet rules) from extracted text content before they enter the build pipeline. - [DATA_EXFILTRATION]: The skill transmits environment-provided authentication tokens (
WIX_AUTH_TOKEN) to official Wix API endpoints (wixapis.com) to provision CMS collections. This transmission is confined to the vendor's own infrastructure. - [EXTERNAL_DOWNLOADS]: The skill downloads dependencies from the official npm registry and browser binaries via Playwright. All external resources are from well-known and established service providers.
Audit Metadata