decklink-docs

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

• Third-party content exposure detected (high risk: 0.85). The skill fetches and text-extracts outsider-authored free-form content from public third-party sources at runtime (e.g., bmdtools-github and GitHub search result pages in PAGES, via load_or_fetch() → fetch_url() → HTML parsing), which can then be included in the agent’s LLM context through search/section outputs.