Shell
/write-research - Generate Research Article Essay
Generate a research article in Eddie's voice for id8labs.app's Research category.
Usage
/write-research "<topic>"
/write-research "Why most productivity tools fail knowledge workers"
/write-research --from-notes <file> # Generate from research notes
When to Use
Research articles are for insights discovered during the building process:
- Market research and competitive analysis
- Technical deep-dives worth sharing
- Patterns observed across users/products
- Philosophical observations about the industry
- "What we learned building X" retrospectives
Process
Gather Research Context
- Topic and core thesis from arguments
- If --from-notes, read the source file
- Identify 3-5 key insights or data points
- Find the deeper pattern (the "why behind the why")
Apply Eddie's Voice Profile
Reference
~/.claude/CLAUDE.md:Research-Specific Moves:
- Pattern recognition across domains (charts to Buddha, fungi to neural networks)
- Etymology and Latin roots for key concepts
- Connect the micro to the macro
- Personal story that illustrates the research
Structure for Research:
- Hook with unexpected observation or personal confession
- "What I was looking for..." (the question that started this)
- Discovery process (share the journey, not just conclusions)
- Pattern connection (how this relates to larger truths)
- Actionable insight (what this means for the reader)
- Open-ended close (invitation to think alongside)
Generate MDX Content
--- title: "[Insight-driven title, not topic-driven]" subtitle: "[The unexpected conclusion or question]" date: "[YYYY-MM-DD]" author: "Eddie Belaval" category: "research" tags: ["research", "{topic}", "{related-concepts}"] ---Content should be 1000-2000 words:
- Personal entry point into the research
- Data/observations (but stories, not just stats)
- Pattern recognition moment
- Connection to reader's experience
- Philosophical or actionable takeaway
Output
- Display full MDX content for review
- Suggest a filename slug
- Offer to proceed with /publish-essay
Voice Calibration for Research
| Do This | Not This |
|---|---|
| "I spent three months in the data and found..." | "Our research indicates..." |
| "Here's what nobody talks about..." | "Industry analysis shows..." |
| "The pattern I keep seeing..." | "Trends suggest..." |
| Connect to personal experience | Pure academic distance |
Example Output
For /write-research "The 70% problem in AI tooling":
---
title: "The 70% Problem"
subtitle: "Why AI tools get us most of the way, then leave us stranded"
date: "2025-12-30"
author: "Eddie Belaval"
category: "research"
tags: ["research", "ai", "productivity", "claude-code"]
---
I've been tracking something for the past six months. A pattern that keeps appearing every time I work with AI tools — and I think it explains why most of us feel like we're living in the future but still working like it's 2019.
Call it the 70% problem.
[Content continues - personal observation, data from experience, pattern recognition]
...
Here's where it gets interesting. The 70% problem isn't a bug. It's a feature of how these systems work. And once you see it...
[Deeper insight, connection to larger pattern]
...
I don't have all the answers here. But I know what questions to ask now.
Maybe that's the real 30%.
Research Tone Markers
- "I've been tracking something..."
- "Here's what nobody talks about..."
- "The pattern I keep seeing..."
- "I spent [time] in the data and found..."
- "Once you see it, you can't unsee it"
- "Maybe that's the real [insight]..."
Requirements
- Eddie's voice profile in ~/.claude/CLAUDE.md
- Clear thesis or observation to explore
- Ideally, some personal experience or data to draw from