workos-authkit-vanilla-js

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

• Third-party content exposure detected (high risk: 0.70). This skill's Step 1 requires a blocking WebFetch of the public GitHub README at https://github.com/workos/authkit-js/blob/main/README.md and declares it the "source of truth," so the agent will read and act on externally-hosted, potentially untrusted third-party content.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill performs a blocking WebFetch of the README at https://github.com/workos/authkit-js/blob/main/README.md and explicitly declares "README is source of truth," so the fetched content is required at runtime and directly controls the agent's instructions.