Prompt: Blog Post Draft
Used by the content-writer agent (Quill) to generate the initial draft.
There are two post types — use the matching template.
Type 1: Solution / Tip Post
For posts that walk through a technical problem and its solution (e.g. Databricks XML ingestion, dbt nested JSON).
Variables
- `` — subject of the post
- `` — intended reader (e.g., data engineer)
- `` — bullet list of points to cover
- `` — post number (e.g., 24)
Prompt
Write a technical solution post about for .
Cover these key points:
Follow this exact structure:
1. Opening (2-3 sentences): Start with a relatable problem or scenario.
Conversational but not fluffy. Reference related past work if relevant.
2. Image embed on its own line:
3. ### Challenges This Solution Addresses:
3-4 bullet points. Each bullet: **Bold Title:** 2-3 sentences of explanation.
Use $~$ between bullets for vertical spacing.
4. ### Solution Logic Explained:
Numbered overview of the full approach — one sentence per step, no code yet.
5. ### Step 1: [Title] ... ### Step N: [Title]
One H3 section per step. Show code first, then explain what it does.
Use language-tagged code blocks (python, sql, bash, etc.).
Include example output tables where helpful.
6. ### Summary
2-3 sentences. Recap what was built and why it matters in production.
7. Sign off with exactly:
Happy data engineering!
Eric
---
*Drafted by **Quill**, reviewed by **Sage** and **Oracle** — the AI agents behind this post.*
Style rules:
- Direct and technical — no filler
- Bold key terms: **Term:** explanation
- $~$ for spacing between bullet points
- Tables for structured data comparisons
Type 2: Opinion / Thought Post
For posts sharing perspective, analysis, or commentary (e.g. “Is Power BI a Self Service Tool?”).
Variables
- `` — the question or thesis
- `` — intended reader
- `` — the key arguments or angles to cover
Prompt
Write an opinion post about for .
Cover these viewpoints and arguments:
Follow this structure:
1. Opening image embed (if available):
2. State the central question or thesis in 1-2 sentences.
3. Explain why this topic matters — what misconception or problem prompted this post?
4. Develop each viewpoint in its own paragraph or short section.
Use --- horizontal rules to separate major shifts in argument.
Use inline images if they illustrate a point.
5. Concrete recommendation or call to action in conclusion.
6. Sign off with exactly:
Thanks
Eric Dong
---
*Drafted by **Quill**, reviewed by **Sage** and **Oracle** — the AI agents behind this post.*
Style rules:
- First person, direct, opinionated — say what you actually think
- No academic hedging ("it could be argued that...")
- Short paragraphs, conversational
- Back opinions with specific observations from real experience