| name | DgBlog |
| description | Blog post authoring for Alternef Digital Garden. USE WHEN writing blog posts, editing posts, EN→FR translation, writing voice questions, moving drafts to published. |
Customization
Before executing, check for user customizations at:
~/.claude/PAI/USER/SKILLCUSTOMIZATIONS/DgBlog/
If this directory exists, load and apply any PREFERENCES.md, configurations, or resources found there. These override default behavior. If the directory does not exist, proceed with skill defaults.
🚨 MANDATORY: Voice Notification (REQUIRED BEFORE ANY ACTION)
You MUST send this notification BEFORE doing anything else when this skill is invoked.
-
Send voice notification:
curl -s -X POST http://localhost:8888/notify \
-H "Content-Type: application/json" \
-d '{"message": "Loading DgBlog skill for blog post authoring"}' \
> /dev/null 2>&1 &
-
Output text notification:
Loading **DgBlog** skill for blog post authoring...
This is not optional. Execute this curl command immediately upon skill invocation.
Digital Garden Blog Authoring
Persistent knowledge layer for blog posts in the Alternef Digital Garden. Covers architecture, frontmatter patterns, content structure, writing voice, and EN→FR translation.
Workflow Routing
| Trigger | Command |
|---|
| Create new blog post | /dg:create (content-type: blog) |
| Edit or improve post | /dg:improve |
| Translate post to French | /dg:translate (uses TranslationWorkflow.md) |
| Publish draft blog post | /dg:publish |
| Sync post to GitHub | /dg:sync |
| Validate blog structure | Add --validate flag to any command |
Context files: BlogStyleGuide.md · TranslationWorkflow.md · BlogPostEnTemplate.md · BlogPostFrTemplate.md
Blog Architecture
content/blog/
├── index.md # Blog landing page
├── the-compiled-stack.md # Technical architecture posts
├── holochain-*.md # Holochain ecosystem posts
├── rea-valueflows-*.md # Economics/protocol posts
├── fractal-sovereignty-*.md # Systems thinking posts
├── fr/ # French translations
│ ├── index.md # French blog landing
│ ├── souverainete-fractale-*.md
│ └── symphonie_des_consciences.md
└── ...
content/unpublished/ # Drafts not yet published
├── index.md
└── *.md # Work-in-progress posts
Frontmatter Pattern
Required fields
---
title: "Post Title: Subtitle with Context"
date: "YYYY-MM-DD"
author: "Soushi888"
description: "Compelling 1-2 sentence summary for SEO and social sharing (~100-200 chars)"
tags:
- "blog"
- "primary-topic"
- "secondary-topic"
- "technology-or-domain"
---
Optional fields
subtitle: "Extended title or hook"
updated: "YYYY-MM-DD"
draft: true
type: "article"
categories:
- "Category Name"
aliases:
- "Short Name"
heroImage: "/blog-placeholder-3.jpg"
language: "fr"
audience: "technical"
Date management rules
date is the publication date — set once when creating the post, never change it afterward.updated is the last modified date — add or update it every time you meaningfully change an existing post.- When creating a new post: set
date only. Do not include updated.
- When editing an existing post: add or update
updated: "YYYY-MM-DD" (today's date).
Author conventions
- Primary author:
"Soushi888" (most posts)
- Organization:
"Alternef" (some older posts)
- AI collaboration:
"Claude & Soushi888" (when AI co-authored)
Tag rules
Tag vocabulary and format standards live in DgTags (TagVocabulary.md). Load DgTags when working with tags. After drafting a post, run SuggestTags from DgTags to get tag suggestions.
Blog-specific exception (stays here): blog is always the first tag for every blog post. DgTags' SuggestTags workflow prepends it automatically for content/blog/ paths.
Post Structure
No H1 title in body
Rule: Never start a post with # Title. The title frontmatter field is the canonical title — Quartz renders it as the page H1. Adding # Title in the body duplicates the heading.
Opening pattern
*Italicized hook restating the core thesis in 1-2 compelling sentences*
---
[Opening paragraphs that set context and frame the problem]
Body organization
- 4-6 major sections (H2
##)
- Each section covers one major concept or technology
- H3 subsections for detailed aspects
- Code examples where relevant (with language specifier)
- Wikilinks to knowledge garden:
[[note-name|Display Text]]
"Honest Tradeoff" pattern
This is a signature style element. After presenting benefits of a technology or approach, include:
**The honest tradeoff:** [Genuine limitations, costs, or challenges with this approach.
Not marketing — real constraints that practitioners should understand.]
This pattern appears in technical posts and demonstrates intellectual honesty. It:
- Acknowledges real costs alongside benefits
- Builds reader trust
- Shows deep understanding of tradeoffs
- Prevents the post from reading like advocacy
Closing pattern
Posts typically end with:
- A synthesis section connecting all threads
- Forward-looking perspective
- Sometimes a call to explore related topics
Images
Required pattern (MANDATORY — no exceptions)
<center>
<img src="filename.jpeg" alt="Descriptive alt text" width="50%" />
</center>
Rules
- Always use the HTML
<img> tag inside <center> — never markdown  syntax
- Filename only — no folder prefix. Quartz resolves images by shortest path match across the content tree. Images live in
content/blog/assets/ but are referenced by name alone
- Always
width="50%"
- Position: after the
--- separator that follows the opening italic hook — never before the hook, never in the frontmatter
Correct position in a post
*Italicized opening hook.*
---
<center>
<img src="image.jpeg" alt="Description" width="50%" />
</center>
First paragraph of body content...
Image storage
Place image files in content/blog/assets/. They resolve correctly from any post location (blog, substack, etc.) via Quartz's shortest-path resolution.
Writing Voice
Characteristics
- Analytical but accessible: Deep technical detail explained clearly
- Honest: Acknowledges limitations alongside strengths
- Concrete-first: Starts with specific examples, generalizes after
- Precise vocabulary: Uses correct technical terms, defines when needed
- No marketing language: Never "blazingly fast", "revolutionary", "game-changing"
- Measured confidence: States what is known vs. what is speculative
Tone markers
- Direct declarative sentences
- "This is..." / "What this means is..." / "The result is..."
- Questions used sparingly to frame sections: "So the question becomes: ..."
- Italics for emphasis and hooks, not bold (bold for technical terms)
Examples from existing posts
Good opening (from The Compiled Stack):
Every layer ships only what it uses. Every tool compiles down to its essence. Nothing carries unnecessary runtime weight.
Good tradeoff (from The Compiled Stack):
The honest tradeoff: Rust's learning curve is real and famously steep. The borrow checker, lifetime annotations, and trait system demand a significant upfront investment...
Good framing:
This is not a sharing economy. It is a micro-service economy with better marketing.
See BlogStyleGuide.md for detailed voice guidance with more examples.
Blog-to-Knowledge Connection
Blog posts should link to knowledge garden notes where relevant:
- Technical concepts:
[[rust|Rust]], [[holochain|Holochain]]
- Theoretical frameworks:
[[rea-accounting|REA Accounting]]
- Governance models:
[[open-value-networks|Open Value Networks]]
This creates bidirectional connections between blog analysis and reference knowledge.
Wikilink rules
When linking to knowledge notes, follow these rules from DgNotes:
- Regular notes:
[[note-filename|Display Name]] (shortest path, no folder prefix needed)
- Index files: ALWAYS use the full absolute path from
knowledge/:
- WRONG:
[[mystical-oriented-programming|...]]
- WRONG:
[[mystical-oriented-programming/index|...]]
- CORRECT:
[[knowledge/tools-and-technology/.../mystical-oriented-programming/index|...]]
- Never use relative paths (
../) — they break in Quartz
Garden MCP Tools — MANDATORY FIRST
BLOCKING RULE: When looking up existing garden content while writing or editing a blog post, a garden MCP tool MUST be called before any Grep/Glob. Do not grep for [[note-name]] patterns or search for existing notes without first querying the index.
| Operation | Primary Tool | Fallback (only if MCP unavailable or stale) |
|---|
| Find knowledge notes on a topic | mcp__garden__garden_search | Grep |
| How a concept connects in the graph | mcp__garden__garden_context | Read + Grep |
| Blog posts already referencing a knowledge note | mcp__garden__garden_backlinks | Grep |
| Explore related concepts before writing | mcp__garden__garden_explore | Multiple Grep |
Index freshness after edits: After saving a blog post, the index updates within ~1-2s. Use Read directly for immediate post-write verification. For bulk blog updates (e.g., fixing frontmatter across multiple posts), call mcp__garden__garden_status to confirm the index is current before running any MCP searches.
French Translation
French translations live in content/blog/fr/. See TranslationWorkflow.md for the complete EN→FR workflow.
Key conventions:
- Add
language: "fr" to frontmatter
- Translate title, description, and content tags
- Keep
"blog" tag untranslated
- File naming: kebab-case in French (e.g.,
souverainete-fractale-integration-multiechelle.md)
Examples
Example 1: Write a new blog post
User: "Write a blog post about my experience with Holochain"
→ Loads DgBlog skill, reads BlogStyleGuide.md for voice
→ Invokes /dg:create with content-type blog
→ Creates with required frontmatter: title, date, author "Soushi888", tags starting with "blog"
→ Opens with italicized hook, includes "honest tradeoff" pattern, links to knowledge notes
Example 2: Translate a post to French
User: "Translate the compiled stack post to French"
→ Loads DgBlog skill, reads TranslationWorkflow.md
→ Invokes /dg:translate
→ Creates file in content/blog/fr/ with language: "fr" frontmatter
→ Translates title, description, and body while keeping code blocks intact
Example 3: Publish a draft post
User: "Publish the shadow work draft"
→ Loads DgBlog skill
→ Invokes /dg:publish
→ Moves from content/unpublished/ to content/blog/
→ Removes draft: true from frontmatter, sets publication date
