Workflow composite that runs manager-portfolio-diff across a curated cohort of well-known filers (Berkshire, Baupost, Renaissance, Bridgewater, Third Point, Pershing Square, Tiger Global, Scion, Appaloosa) and aggregates the initiations, adds, and exits by issuer. Surfaces cross-fund conviction: names that appeared in >= N funds' new positions this quarter. Requires Stocks Basic. Runs on the free tier.
Installs into .claude/skills of the current project.
Are you the author of Smart Money Cluster?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/rgourley-smart-money-cluster)
---
name: smart-money-cluster
description: Workflow composite that runs manager-portfolio-diff across a curated cohort of well-known filers (Berkshire, Baupost, Renaissance, Bridgewater, Third Point, Pershing Square, Tiger Global, Scion, Appaloosa) and aggregates the initiations, adds, and exits by issuer. Surfaces cross-fund conviction: names that appeared in >= N funds' new positions this quarter. Requires Stocks Basic. Runs on the free tier.
---
# smart-money-cluster
Runs `manager-portfolio-diff` across a curated cohort of well-known
funds and aggregates the results by issuer. Surfaces names that
appeared in >= N funds' new positions (initiations), adds, or exits
this quarter as a "cross-fund conviction" signal.
Default cohort: Berkshire, Baupost, Renaissance, Bridgewater, Third
Point, Pershing Square, Tiger Global, Scion, Appaloosa. Custom
cohort via `--aliases`.
## When to invoke
- Quarterly 13-F review
- "What did smart money buy this quarter?"
- Screening for names with cross-fund conviction as a starting point
- The user says "smart money", "13-F cluster", "cross-fund"
Not for: real-time (13-F is quarterly and lagged ~45 days). Not for
alpha timing (crowded positions can underperform).
## What you need
- `MASSIVE_API_KEY` exported
- Stocks Basic minimum
Optional:
- `--aliases` (default: 9-fund cohort)
- `--min-funds` (default 2): minimum fund count to surface a name
## What you get back
**Layer 1: JSON**. Per-fund summary, clustered_initiations,
clustered_adds, clustered_exits, each sorted by fund count then
dollar volume.
**Layer 2: rendered note**. Per-fund one-liners + three cluster
tables + Take.
## Foundations used
- Composes `manager-portfolio-diff` across N filers.
## Doesn't handle (yet)
- **Custom cohort scoring.** Each fund weighted equally. Historical
accuracy weighting would be a real extension.
- **Price context.** No overlay of cluster picks vs current price /
RS / vol regime.
- **Time-series cluster tracking.** Only current quarter; no
quarter-over-quarter momentum.
These are clean composite extensions.
