sdk-ut-boundary-generator

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

• Third-party content exposure detected (medium risk: 0.30). required runtime ingests outsider-authored free text from user-provided remote repository content (e.g., README/docs files selected by the input URL) by cloning the specified GitCode/GitHub/GitLab repo and then reading docs/**/*.md to extract API boundary definitions.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly performs runtime cloning of user-provided remote repositories (e.g., runs "git clone --depth 1 https://gitcode.com/owner/repo.git") and then reads API docs from those repositories to drive prompt/content used to generate tests, so external URLs like https://gitcode.com/owner/repo.git are fetched at runtime and directly control the agent's outputs.