sentiment-report

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

• Third-party content exposure detected (high risk: 1.00). This skill explicitly fetches and ingests public, user-generated content from news sites, Reddit, and StockTwits (see Step 1 WebSearch, Step 2 invoking reddit-sentiment/stocktwits-sentiment, and the "Data Source Log" referencing NewsAPI/Reddit/StockTwits), and it directly uses that untrusted content to compute composite scores and drive reporting, which could enable indirect prompt injection.