Mode 2: From Bullet Points / Description

User provides brief outline, bullet points, or topic description without file path.

/zola-blog
- topic: setting up WezTerm for AI coding
- pain point: opening multiple terminals every morning
- solution: Lua script to auto-launch agents
- features: hotkey launcher, auto-maximize, tab management

Detect mode by checking if user input contains a valid file path (.md extension with drive letter or starts with path separator).

Workflow for Mode 1 (From File)

Step 1: Read Source File

Read the Obsidian markdown file from the provided path.

Step 2: Convert Obsidian to Zola Format

Use the existing import_markdown function from the toolkit to handle:

Remove Obsidian YAML frontmatter
Convert wikilinks [[link]] to plain text
Remove callout formatting > [!note]
Process images: copy from Obsidian vault to blog static folder

Call the Python toolkit:

cd "D:\pcloud\workspace\code\python"
python -c "from toolkit_modules.tools.blog import BlogTools; BlogTools().import_markdown()"

Or use the conversion functions directly if needed.

Step 3: Translate and Optimize

After conversion, translate and optimize the content:

Translate Vietnamese to English
- Keep technical terms in English
- Maintain code blocks unchanged
- Preserve markdown formatting
Optimize writing style - Follow references/blog-guidelines.md strictly:
- Remove AI buzzwords completely
- Use personal voice: "I use", "I found", "I built"
- Short paragraphs (2-4 sentences max)
- Specific numbers instead of vague claims
- Honest about pros AND cons
- Active voice only
Expand content where needed
- Add real examples from context
- Include "who should use this" section
- Add comparison with alternatives if relevant

Step 4: Review Checklist

Before publishing, verify:

No AI buzzwords (seamlessly, powerful, leverage, game-changer, etc.)
Personal voice throughout ("I", not "we" or passive)
Opening hook is specific pain point, not generic
All numbers are concrete, not vague
Short paragraphs (2-4 sentences)
Reads like human developer wrote it

Step 5: Write Final Post

Write the optimized English content to the blog:

Location: D:\pcloud\workspace\code\website\quoc app\content\blog\
Filename: slug from original filename
Include Zola frontmatter with title, date, tags

Step 6: Build and Deploy

Provide commands for user to run:

cd "D:\pcloud\workspace\code\website\quoc app"
zola build
git add .
git commit -m "Add new blog post: [title]"
git push

Workflow for Mode 2 (From Bullet Points / Description)

Step 1: Understand the Input

User provides outline in Vietnamese or English:

Topic / subject
Pain point or problem
Solution or approach
Key features or points to cover
Personal experiences (if any)

Example input:

/zola-blog
- chủ đề: dùng marker convert PDF sang markdown
- vấn đề: PDF khó edit, copy text bị lỗi format
- giải pháp: dùng marker, hỗ trợ OCR, giữ được format
- tính năng: CLI đơn giản, hỗ trợ batch, extract ảnh

Step 2: Research and Expand

Before writing, gather more context:

Ask clarifying questions if needed (target audience, technical depth, specific examples)
If topic is technical, understand the tool/concept fully
Identify real use cases and scenarios

Step 3: Write Full Blog Post

Write complete blog post in English following references/blog-guidelines.md:

Structure to follow:

Opening hook - specific pain point, personal story
Problem section - why existing solutions fail
Solution introduction - what it does, no hype
Features breakdown - each with real example
Real workflow examples - step-by-step with numbers
Comparison with alternatives - honest pros/cons
Who should use this - honest assessment
Setup guide (if technical)
Bottom line - restate benefit, simple CTA

Critical rules:

Write as if YOU (the author) experienced this
Use "I found", "I use", "I tried"
Include specific numbers: "takes 30 seconds", "saved 2 hours"
Be honest about limitations
NO AI buzzwords

Step 4: Review Checklist

Same as Mode 1 - verify:

Step 5: Generate Slug and Write

Generate slug from topic: marker-pdf-to-markdown
Write to blog content folder
Include proper Zola frontmatter

Step 6: Build and Deploy

Same as Mode 1 - provide git commands.

Key Paths

Obsidian vault: D:\pcloud\workspace\obsidian\
Obsidian images: D:\pcloud\workspace\obsidian\90 luu tru\images\
Blog root: D:\pcloud\workspace\code\website\quoc app\
Blog content: D:\pcloud\workspace\code\website\quoc app\content\blog\
Blog images: D:\pcloud\workspace\code\website\quoc app\static\images\blog\

Writing Guidelines Reference

For detailed writing guidelines, anti-AI patterns, and examples, see references/blog-guidelines.md.

Example Usage

Mode 1: From File

User: /zola-blog D:\pcloud\workspace\obsidian\10 cong cu\neovim\lazyvim.md

Agent will:

Read the Vietnamese source file
Convert Obsidian syntax to standard markdown
Copy any referenced images to blog static folder
Translate to English
Optimize writing following guidelines
Write to blog content folder
Provide git commands to publish

Mode 2: From Outline

User: /zola-blog
- topic: using Marker to convert PDF to markdown
- problem: PDFs are hard to edit, copy-paste breaks formatting
- solution: Marker preserves structure, supports OCR
- features: simple CLI, batch processing, image extraction
- my experience: converted 50 research papers last week

Agent will:

Understand the outline and ask clarifying questions if needed
Research the topic for accuracy
Write complete blog post in English with personal voice
Follow anti-AI-detection guidelines strictly
Generate appropriate slug and frontmatter
Write to blog content folder
Provide git commands to publish

zola-blog

Zola Blog Publisher

Two Modes

Mode 1: From Obsidian File