| name | transcript-analyze |
| description | Deep thematic analysis of specific transcripts or cross-cutting topics across the library |
| tags | ["transcript","analyze","theme","pattern","deep-dive","nate-jones"] |
| version | 1.0.0 |
/transcript-analyze — Deep Transcript Analysis
Thematic deep-dive into one or more transcripts. Finds patterns, contradictions, and evolving positions.
When to use
- "Analyze [specific video title]"
- "How has Nate's position on [topic] changed over time?"
- "What patterns emerge across the [cost / coding / market] videos?"
- "Compare what Nate said about X in February vs March"
- "Find contradictions in the transcript library"
Process
- Check Chartroom for prior analysis on the topic
- Pull relevant transcripts via Research agent's transcript-query
- For single video: extract thesis, evidence, predictions, action items
- For cross-cutting: map theme evolution across dates, find turning points
- Tag findings: LEVERAGE, SUSTAIN, or EDUCATE
- Flag predictions that are verifiable — note verification date
Output Format
## Transcript Analysis: [Topic/Title]
**Scope**: [N videos, date range]
**Thesis**: [Core argument in 1-2 sentences]
### Key Points
1. [Point] — [video, date]
### Evolution (if cross-cutting)
| Date | Position | Video |
|------|----------|-------|
### Predictions (verifiable)
| Claim | Verify By | Status |
|-------|-----------|--------|
### Strategic Takeaway
[What this means for us — tagged LEVERAGE/SUSTAIN/EDUCATE]
Rules
- Quote directly when the original wording matters
- Flag contradictions explicitly — they're valuable signal
- Date everything — positions shift fast in AI
- If analysis needs more transcripts ingested, request via Research
Intent: Informed [I18]. Purpose: Deep intelligence extraction.