research

name: research description: Use when gathering sources and background material for a newsletter article topic. Takes a topic as argument.

Gather and structure source material for an article topic.

Arguments

/research <topic> — e.g., /research AI agents in enterprise sales

Process

Search broadly — Use WebSearch to find 8-12 sources on the topic. Prioritize:
- Contrarian or novel angles (not the obvious take)
- Data points, statistics, and named examples
- Cultural or philosophical references that fit Chris's style
- Primary sources over aggregators
Deep-read the best sources — Use WebFetch on the top 5-6 results. Extract:
- Key claims and supporting evidence
- Specific company names, numbers, and quotes
- Tensions or contradictions between sources
- What conventional wisdom gets wrong
Structure the notes — Save to research/YYYY-MM-DD-<topic-slug>.md with this format:

---
topic: "<topic>"
date: YYYY-MM-DD
sources: []
---

## Key Findings

<3-5 bullet summaries of the most important insights>

## Contrarian Angles

<What does the mainstream narrative miss? Where is conventional wisdom incomplete?>

## Data Points & Examples

<Specific numbers, company names, quotes — the concrete material for the article>

## Tensions & Open Questions

<Contradictions between sources, unresolved debates, areas needing more depth>

## Sources

<Numbered list of URLs with one-line descriptions>

Summarize — Present a brief summary of findings to the user, highlighting the strongest contrarian angle and most compelling data points.

Key Principles

Prioritize depth over breadth — 5 well-read sources beat 15 skimmed ones
Look for the "everyone says X, but actually Y" angle — this drives Chris's best articles
Capture specific names and numbers — vague trends don't make good openings
Note cross-domain analogies — Chris uses these for intellectual texture