dual-listed-arbitrage

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

• Third-party content exposure detected (medium risk: 0.30). Runtime path ingests outsider-authored free text: the skill takes the user’s natural-language prompt and, when it contains symbols (e.g., CHKP/NICE/TEVA), those are used to fetch US/TASE price data via scripts/provider.py (yfinance) and to render pair labels in scripts/viz.py; there is no requirement that the user select a specific pre-registered item before ingestion/processing.