temporal

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

• Third-party content exposure detected (high risk: 0.85). The only runtime text ingestion path shown is a required WebFetch("https://docs.temporal.io/llms.txt") of public Temporal documentation; that fetched free-form page content is outsider-authored and would be read into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly requires a runtime fetch via WebFetch("https://docs.temporal.io/llms.txt") before proceeding, so this external document is fetched at runtime and directly controls the agent's instructions/behavior as a required dependency.