Ai-first-toolkit write-blog-post

Write or develop a blog post. Use for blog content - writing, drafting, developing ideas into drafts, or editing. Longer-form than LinkedIn (800-1200 words) with section structure.

install
source · Clone the upstream repo
git clone https://github.com/techwolf-ai/ai-first-toolkit
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/techwolf-ai/ai-first-toolkit "$T" && mkdir -p ~/.claude/skills && cp -r "$T/plugins/content-studio/skills/write-blog-post" ~/.claude/skills/techwolf-ai-ai-first-toolkit-write-blog-post && rm -rf "$T"
manifest: plugins/content-studio/skills/write-blog-post/SKILL.md
source content

Write Blog Post

You are helping write a blog post for the author.

Before Writing (MANDATORY)

You MUST complete these steps before writing ANY content:

  1. Run
    ./scripts/print-published.sh linkedin-post
    to read ALL published LinkedIn posts
    • Blog posts share the same voice as LinkedIn posts, so these are your primary style reference
    • Note core insights and data points already used
    • Identify opportunities to reference or build on previous posts
  2. Run
    ./scripts/print-published.sh blog-post
    to read ALL published blog posts
    • Avoid repeating topics, angles, or arguments already covered
    • Note the structure and depth of existing blog posts
  3. Read
    guidelines/linkedin.md
    for style rules (same voice applies to blog posts)
  4. Read
    references/professional-profile.md
    for background

If developing an idea-stage post: Check if the idea's core insight overlaps with published posts (LinkedIn or blog). If so, either:

  • Find a genuinely different angle
  • Explicitly build on the previous piece
  • Recommend against developing the idea

Avoid Repetitive Patterns

When reading recent posts, actively note and vary:

Openings: Vary between demand signal, personal anecdote, company experience, surprising data Section flow: Don't always follow the same arc (problem-solution-data-close) Closing lines: Find fresh ways to land the argument Examples: Rotate between company-specific, industry, and broader examples Rhetorical devices: If recent pieces use lists heavily, try flowing narrative, and vice versa

The goal is a consistent voice with varied execution.

Style Requirements

  • 800-1200 words (use
    wc -w
    to verify)
  • English language
  • Section headings (## level) to break up the piece into 5-7 sections
  • First person throughout
  • Conversational tone - read it aloud, does it sound like something you'd say?
  • Specific > abstract - lead with concrete examples, generalize later
  • No hype words ("revolutionary", "game-changing", "transformative")
  • Strong opening - first 2-3 sentences must hook the reader with something concrete

Structure

Blog posts are longer than LinkedIn posts and need more structure. A typical blog post has:

  1. Hook (80-100 words) - Concrete opening that pulls the reader in
  2. 3-5 body sections (150-200 words each) - Each with a clear heading and one main point
  3. Close (50-75 words) - Echo the opening or land the core argument

Each section should earn its place. If a section doesn't add something the reader couldn't get from the LinkedIn post on the same topic, cut it.

Relationship to LinkedIn Posts

Blog posts often expand on ideas first introduced in LinkedIn posts. When this happens:

  • Reference the LinkedIn post briefly, don't re-explain it
  • Go deeper on the angle the LinkedIn post couldn't cover in 200 words
  • Add the meta-story, the data, or the "why" behind the insight
  • Blog posts should reward readers who already saw the LinkedIn post AND stand alone for those who didn't

Process

  1. If given a topic, develop it into a full draft
  2. If given an existing idea file (search with
    ./scripts/search-posts.sh
    ), develop the content
  3. Always check word count against 800-1200 target
  4. Save to content/posts/blog-post/ with type: blog-post, stage: 02-drafts

Validation

After saving the draft, run the word count checker:

scripts/check-word-count.sh <yaml-file>

Adjust the content if the word count is outside the 800-1200 target.

Sub-Agent Review

Before presenting the final draft, spawn a sub-agent to review with fresh eyes. The reviewer should:

  • Check voice consistency against
    guidelines/linkedin.md
    (same voice applies)
  • Verify each section earns its place and adds depth beyond a LinkedIn post
  • Flag any overlap with published posts (LinkedIn or blog)
  • Check for hype words, generic phrasing, or sections that could be cut

Incorporate the reviewer's feedback before finalizing.

Creating New Files

Get timestamp first:

date -u +"%Y%m%d-%H%M%S"  # For slug
date -u +"%Y-%m-%dT%H:%M:%S.000Z"  # For created/lastUpdated

File path:

content/posts/blog-post/{slug}-{slugified-title}.yaml