add-feature

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly instructs the agent to fetch and follow raw GitHub doc URLs and an external public gist (e.g., the ShadCN guide at https://gist.githubusercontent.com/infomiho/.../raw/...), meaning it will read and act on open, third‑party public content that could be untrusted.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs at runtime to fetch and follow the external raw Gist guide (https://gist.githubusercontent.com/infomiho/b35e9366e16913949e13eaba0538f553/raw/c6da98158c1a7e46b5874868f2e7c011f24d24d1/0-README.md), which directly controls the agent's installation/configuration instructions and is used as a required dependency for the ShadCN UI setup, so it meets the criteria for a runtime-controlled external prompt/code dependency.