core-animation

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

• Potentially malicious external URL detected (high risk: 0.80). The skill's "Fetching More Docs" instructions explicitly direct runtime fetching from the sosumi.ai Markdown mirror (e.g. https://sosumi.ai/documentation/quartzcore/camediatimingfunction), which would be injected into the agent's context to answer queries and thus can directly control prompts.