content-hydration

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

• Third-party content exposure detected (high risk: 0.95). The skill’s runtime path fetches Reddit permalinks as JSON via the “.json trick” (requests.get(url + ".json") in fetch_reddit_json), and the fetched post/comment text is then parsed and joined into enriched CSVs—this is outsider-authored content from Reddit users (not chosen by the operating user as input prompts).