clonelisted

# Clone — superinvestor 13F sweep Load the firm-doctrine skill (philosophy: cloning sources ideas, never conclusions). ## Procedure 1. Dispatch **idea-scout** with the objective: sweep `get_13f_holdings` for the standard roster (berkshire, himalaya, baupost, pabrai, oaktree, aquamarine, tweedy, fairfax) — plus any manager the user names (by CIK if not in the roster). 2. Instruct it to focus on: NEW positions, large adds (>25% share increase), and concentrated high-conviction weights (>5% of a manager's 13F) — and to note for each: manager, weight, estimated cost context (price range during the quarter via `get_prices`), and whether the current price is above or below that range. 3. Require the caveats in the output: 45-day lag; US-listed longs only; a position may already be sold; Himalaya's non-US book is invisible; we don't know their reasons. 4. On return, apply the firm filter inline: circle of competence, plausible cheapness TODAY (not at their cost), analyzability. Route each survivor: TRIAGE candidate / ignore. ## Output - Table of noteworthy 13F moves with source tags (accession numbers) - Which 2–3 names deserve triage and the one-line reason each - The reminder, verbatim: "A cloned idea still goes through the full process — triage, deep dive, red team, checklist. We copy homework sources, not homework answers."