Peter Yang's AI Content System

<task>
Research [Guest Name] for upcoming interview.
</task>

<sources>
- Twitter/X: @handle
- LinkedIn profile
- Recent blog posts
- Podcast appearances
- GitHub projects
</sources>

<output>
1. Key themes in their work
2. Unique perspectives or contrarian views
3. Tools/workflows they've mentioned
4. 10 specific questions based on their public content
5. Gaps in what's publicly known (areas to explore)
</output>

<raw_transcript>
{{transcript}}
</raw_transcript>

<task>
Extract publishable content from this interview.
</task>

<output_format>
## Key Quotes (verbatim, with timestamps)
## Workflow Descriptions (step-by-step)
## Tools Mentioned (with context)
## Unique Insights (things not commonly known)
## Follow-up Questions (for clarification or Part 2)
</output_format>

<role>
Tech content editor writing for product managers and
developers interested in AI workflows.
</role>

<source_material>
{{processed_interview_notes}}
</source_material>

<article_structure>
1. Hook: One surprising insight from the interview
2. Context: Who is this person, why listen to them
3. System Overview: High-level workflow
4. Deep Dives: 3-4 specific techniques with examples
5. Getting Started: How readers can apply this
6. Resources: Links to tools, further reading
</article_structure>

<constraints>
- 1500-2500 words
- Include code blocks where applicable
- Use tables for comparisons
- End with actionable next steps
</constraints>

<article>
{{full_article}}
</article>

<task>
Create a Twitter/X thread from this article.
</task>

<constraints>
- 8-12 tweets
- First tweet: hook with surprising insight
- Last tweet: call to action + link
- Each tweet: one complete thought
- Include relevant images/screenshots where applicable
- Use line breaks for readability
</constraints>

<task>
Write the "Featured Interview" section for this week's newsletter.
</task>

<format>
## This Week's Interview: {{guest_name}}

[One-sentence hook]

**The Big Insight:**
[2-3 sentences on the key takeaway]

**One Thing to Try:**
[Specific action reader can take today]

[Read the full interview →](link)
</format>

<product_context>
{{company_description}}
{{existing_features}}
{{target_users}}
</product_context>

<feature_request>
{{feature_description}}
</feature_request>

<output_sections>
1. Problem Statement
2. User Stories
3. Success Metrics
4. Requirements (functional)
5. Requirements (non-functional)
6. Out of Scope
7. Open Questions
</output_sections>

<style_guidelines>
- Specific over vague
- Measurable success criteria
- Clear scope boundaries
- Acknowledge unknowns explicitly
</style_guidelines>

<role>
Senior product manager reviewing a PRD draft.
Be constructive but thorough.
</role>

<prd>
{{draft_prd}}
</prd>

<review_criteria>
- Clarity: Can engineering build from this?
- Completeness: Are edge cases addressed?
- Feasibility: Is scope realistic?
- Metrics: Are success criteria measurable?
- Risks: Are dependencies and blockers identified?
</review_criteria>

<output>
## Strengths
## Areas for Improvement
## Missing Elements
## Questions for Clarification
## Suggested Next Steps
</output>

<current_pipeline>
{{pipeline_status}}
</current_pipeline>

<task>
Review my content pipeline and identify:
1. What's ready to publish this week
2. What needs work
3. What's falling behind
4. Gaps in my content mix
</task>

<content_goals>
- 2 long-form articles per week
- 1 video episode per week
- Daily social media presence
- Weekly newsletter
</content_goals>

Track for one week:
- What content do you create?
- How long does each piece take?
- What's repetitive?
- What's bottlenecking production?

Example: Weekly newsletter

Template:
1. Intro hook (AI: generate from this week's theme)
2. Main content (Human: core insight)
3. Supporting points (AI: expand on outline)
4. CTA (Template: reuse)
5. Links (Human: curate)

Content Pipeline:
Ideas → Research → Draft → Edit → Publish → Distribute

For each stage:
- What can AI assist with?
- What needs human judgment?
- What's the handoff process?

For each piece of content:
- What other formats can this become?
- What platforms should it reach?
- How can AI help with reformatting?