us-screenerlisted

# US Screener Skill ## When to use - "Run the screener" / "Find candidates" / "What's setting up?" - Weekly or daily morning routine to surface new names matching doctrine - Before adding any new US name to watchlist — verify it passes P1 + has reasonable quality/value - "Show me Buffett candidates" / "Anything cheap and trending?" Do NOT use for: - Crypto discovery (use `crypto-coingecko` + future crypto-screener) - KLSE discovery (no equivalent yet — small market, manual curation works) - Existing watchlist analysis (use dashboard's per-ticker grids) ## Why this exists The app was strong on analysis but weak on discovery. Manually scanning hundreds of names for P1 setups is a poor use of time. This skill automates the universe scan and only surfaces names that: 1. Pass all 8 P1 technical gates (same logic as `us_status()` in dashboard.py) 2. Aren't already on the watchlist (no noise on names you already track) 3. Carry a Buffett tag showing whether they're quality compounders, value plays, or technical-only The output answers: "Which names should I consider adding to my watchlist today?" ## Universe Hardcoded JSON at `.claude/skills/us-screener/universe.json` — ~180 liquid US names across all 11 sectors. Curated, not the full Russell 1000, because: - Smaller universe = faster scans (~30-60s instead of 5+ minutes) - Liquidity-filtered (no microcaps where the doctrine doesn't fit) - Editable: add names you'd consider trading, remove ones you wouldn't Refresh cadence