keenable-web-search

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow runs scripts/search.py, which POSTs a user-supplied natural-language query to the outsider-authored web via https://api.keenable.ai/v1/search/public and ingests the returned result titles/snippets as LLM-readable text (even though it doesn’t fetch arbitrary pages, it still consumes outsider-provided snippets).