searching-mlflow-docs

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

• Potentially malicious external URL detected (high risk: 0.90). The skill runtime explicitly fetches https://mlflow.org/docs/latest/llms.txt and then https://mlflow.org/docs/latest/[path].md to locate and inject documentation (including verbatim code blocks) into the agent context, meaning external content directly controls what is presented to the model and is a required runtime dependency.