thesis-cluster-detector

Cluster news items by investment thesis using embedding similarity or keyword co-occurrence.

4|1|Updated Apr 11, 2026
One-click install
npx skills add https://github.com/m2ai-portfolio/m2ai-skills-pack --skill thesis-cluster-detector
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: thesis-cluster-detector
Source: https://github.com/m2ai-portfolio/m2ai-skills-pack/tree/main/skills/thesis-cluster-detector
Command: npx skills add https://github.com/m2ai-portfolio/m2ai-skills-pack --skill thesis-cluster-detector

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ingests a batch of enterprise AI or technology news items (from RSS, Substack previews, or pasted headlines) and clusters them by underlying investment thesis using embedding similarity or keyword co-occurrence, then emits a "what is the underlying bet" briefing when a cluster reaches a configurable threshold. Use when the user says "thesis cluster", "what's the underlying bet here", "cluster these stories", "detect the pattern", or pastes multiple news items and wants to know if they are separate events or a single coordinated bet.

Core Features & Use Cases

  • Thesis detection: cluster multiple items to surface shared investment theses.
  • Mode versatility: supports interactive pastes, batch file processing, or scheduled feed scanning.
  • Briefing output: emits a deterministic one- or multi-cluster briefing when thresholds are met, plus singleton notes for isolated items.

Quick Start

Provide a batch of headlines or summaries, and the system will cluster them by underlying investment thesis.

Frequently Asked Questions about thesis-cluster-detector

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I cluster news headlines by underlying investment thesis?

To detect investment theses from pasted news items, provide a batch of headlines or summaries to the system. It groups them by embedding similarity or keyword co-occurrence and emits a briefing when a cluster reaches your configured threshold.

What is the best way to find shared investment bets across multiple news stories?

The best way to find shared investment bets is to process a batch of news items through thesis clustering. It identifies whether separate events form a single coordinated bet and outputs a one-sentence thesis per cluster with singleton notes for unclustered items.

Can I process scheduled RSS feeds to detect emerging investment patterns?

Yes, you can process scheduled RSS feeds to detect emerging investment patterns. The system supports interactive pastes, batch file processing, and scheduled feed scanning, emitting a briefing only when a cluster reaches the configured threshold.

Does news clustering work with batch files and Substack previews?

Yes, news clustering works with batch files and Substack previews. It ingests technology news from multiple sources and applies embedding similarity or keyword co-occurrence to group items into coordinated investment bets.

How do I handle unclustered items when grouping news by investment thesis?

When grouping news by investment thesis, unclustered items are handled as singleton notes. The deterministic cluster report isolates these items so you can review isolated events separately from the detected coordinated investment bets.