xerj-code

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

• Third-party content exposure detected (medium risk: 0.30). The workflow clones reference repositories from GitHub (github.com/...) and indexes their source files into a local search server, which the agent subsequently queries to retrieve reference passages before writing code.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs cloning git repositories from GitHub during corpus setup, but these URLs reference arbitrary project repositories rather than recognized official publisher locations, constituting an unknown remote external dependency.