pull-skills

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

• Third-party content exposure detected (high risk: 1.00). The skill's workflow explicitly runs a git pull on the goatedskills GitHub repo (itsAR-VR/goatedskills) and installs those repo files into local agent skill directories, meaning the agent will ingest third-party repository content that can change its behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). This skill's runtime workflow runs a git pull to fetch the goatedskills repository (https://github.com/itsAR-VR/goatedskills), and the fetched files are distributed into local agent skill directories where they directly control agent prompts/behavior.