thatopen-core-fragments

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). The skill's required workflow calls fragments.init(workerURL) and examples explicitly show loading a worker from a public CDN (e.g., "https://unpkg.com/@thatopen/fragments@3.3.6/dist/Worker/worker.mjs" in SKILL.md and references/examples.md), meaning the agent will fetch and execute code from an arbitrary public URL which can materially alter behavior.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill calls fragments.init at runtime with the remote worker URL "https://unpkg.com/@thatopen/fragments@3.3.6/dist/Worker/worker.mjs", which downloads and executes a remote web-worker JS module that is required for core functionality.