LinkedIn Writer Skill for Claude Code

A Claude Code skill that writes LinkedIn posts in your distinctive voice and style, not generic AI slop.

What It Does

• Accepts any input: rough notes, bullet points, article summaries, or just a topic
• Produces posts that follow a proven structure: Hook > Problem > Named Framework > Action > CTA
• Applies anti-slop rules to eliminate AI writing patterns
• Enforces LinkedIn’s 3,000 character limit automatically
• Supports 5 content pillars with distinct tone guidance

Content Pillars

Installation

Option 1: Clone to your Claude Code skills directory

git clone https://github.com/kvsdileep/linkedin-writer.git ~/.claude/skills/linkedin-writer

Option 2: Manual setup

1. Create the skill directory: mkdir -p ~/.claude/skills/linkedin-writer
2. Copy all files from this repo into that directory
3. Ensure the structure looks like:

~/.claude/skills/linkedin-writer/
  SKILL.md
  references/
    content-pillars.md
    hooks-framework.md
    sample-posts.md
    style-guide.md

Usage

In Claude Code, just ask it to write a LinkedIn post:

Write a LinkedIn post about [your topic]

Or use the slash command:

/linkedin-writer [your notes or topic]

Customization

To make this skill match YOUR voice:

1. references/sample-posts.md - Replace the 5 sample posts with your own best-performing LinkedIn posts. These calibrate the tone.
2. references/style-guide.md - Adjust banned phrases, vocabulary patterns, and signature transitions to match your writing style.
3. references/content-pillars.md - Redefine the 5 pillars to match your content strategy.
4. references/hooks-framework.md - The hook formula works universally, but you can add your own hook examples.

Key Features

• 3-Step Hook Formula: Context Lean-In > Scroll Stop Interjection > Contrarian Snapback
• Named Frameworks: Every post distills insights into a memorable, titled framework
• Character Limit Enforcement: Stays within LinkedIn’s 3,000 character limit with optimal range targeting (1,800-2,800)
• Anti-Slop Engine: Actively scans for and removes AI writing patterns, banned phrases, and structural tells
• CTA Rotation: Engagement questions, repost prompts, and follow prompts

File Structure

License

MIT