heurist-finance: Skill for Claude Code

Small-Cap Scout

Shared plumbing and output contract: ../_shared/analyst-core.md. Small caps are where retail's structural advantage lives (institutions can't buy what won't absorb their size) and where retail gets hunted (illiquidity, promotion, information vacuum). The skill is running the tradeability and integrity gates before falling in love with the story — in that order, because the story is always good; that's what promotions are for.

Gate 1 — Tradeability arithmetic (before any analysis)

From quote_snapshot + prices SYM --period 6mo --full --out FILE compute average daily dollar volume (ADV) and typical spread:

• Spread tax: names under ~$1M ADV cost 1–3% round trip in spread and slippage. Annualized over the user's multi-week turnover, any thesis must clear that hurdle before having edge. State the number.
• Position-vs-ADV rule: intended position ≤ ~1–5% of ADV, sized so a 3-day exit without moving price >1–2% is realistic. The Amihud illiquidity premium (~0.5–1%/month in-sample) is real but non-linear — the bottom liquidity decile is where the premium lives and where retail size can't actually enter and exit; owning it on paper and capturing it are different things.
• Exit stress test: what does getting out look like on the worst day (halt reopens, guidance cut)? If the answer is "I'm trapped," the position size is the problem even when the thesis is right.

Gate 2 — Integrity screen (is this a business or a promotion?)

The microcap manipulation profile is well-defined — check all of these (equity_overview, insider_activity, filing_timeline, web for promotion history):

• The pump signature: < $300M cap + < $5M daily volume + recent +100% move on no filing-linked news + paid-promotion trail (stock "newsletters," X spam) → halt/manipulation risk dominates any thesis; decline the name and say why.
• Dilution machine check: history of shelf offerings, convertibles, warrants (S-1/S-3/424B in filing_timeline; share count trend via xbrl_fact_trends). A company that funds operations by printing shares transfers your upside to itself — cash runway < 4 quarters makes the next offering your event risk.
• Reverse-split history, auditor tier, going-concern language, insider selling into strength, and float held by insiders (small real float = squeeze/rug both ways).
• Sub-$50M-float runners after +100%: you are the exit liquidity for insiders; hard pass regardless of narrative.