write-coding-standards-from-file

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

• Third-party content exposure detected (high risk: 0.85). The skill can run #fetch ${item} for many public URLs when fetchStyleURL == true, ingesting fetched web page text into the LLM context (outsider-authored free text via runtime web content).

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

• Potentially malicious external URL detected (high risk: 0.90). The skill sets fetchStyleURL = true and explicitly instructs the agent at runtime to run "#fetch (URL)" for each link in the "Fetch Links" list (e.g., https://peps.python.org/pep-0008/), using the fetched content as context to compose and control the coding-standards prompts, so external URLs are fetched at runtime and directly influence agent instructions.