One question I keep getting: "How do I know my data is actually private?"

Here's the full picture; it's enforced at three layers, not just one:

Layer 1: Two-tier storage architecture

Your data is split into two buckets in chrome.storage.local:

Currently live: US (NASDAQ/NYSE), Canada (TSX), and India (NSE/BSE). I'm looking at adding LSE (London), ASX (Australia), Shanghai (SSE), and Hong Kong (HKEX) next. Any other markets that would be useful for you? The main constraint is data source availability; yfinance coverage varies by exchange.

StockAgent currently scores stocks on 5 factors: Debt-to-Equity, PEG, ROE, 200-day SMA, and RSI, all with sector-aware thresholds. Thinking about what else could be added while keeping everything deterministic (no AI/ML in the score). Some ideas I'm considering: insider trading activity, earnings surprise history, dividend consistency. What would be most useful to you?

A deliberate design choice in StockAgent: the lowest grade is "AVOID", not "SELL." Saying "sell" is a direct financial instruction that implies you should liquidate a position; that's liability territory. "AVOID" means the rule-based scorecard flagged problems, without prescribing action. Curious if others have opinions on this framing.