linkedin

SKILL.md

LinkedIn Community Engagement Skill

Overview

This skill enables authentic LinkedIn engagement by discovering relevant posts, identifying thought leaders to follow, crafting professional responses, and tracking engagement history.

Actions

Action File When to Use
Login references/login.md Authenticate with LinkedIn before engaging
Research references/research.md Find posts, discover people, build content database
Engage references/engage.md Comment on posts or create new posts

Data Files

File Purpose
.business_growth/marketing/social/BUSINESS.md Shared business context (required)
.business_growth/marketing/social/linkedin/strategy.md LinkedIn engagement strategy (overall approach, target audiences, content pillars)
.business_growth/marketing/social/linkedin/{YYYY-MM}/{YYYY-MM-DD}.md Daily posts, people, and engagement log
community-pulse/skills/linkedin/.env.profiles.local LinkedIn credentials (gitignored)

Quick Start

  1. First time? → Run Login to authenticate, then Research to find posts and people
  2. Have posts? → Run Engage to comment or create new posts
  3. Before any Research or Engagement → The strategy file at .business_growth/marketing/social/linkedin/strategy.md is loaded automatically. If it doesn't exist, you'll be asked if you want to create one.

Decision Flow

User Request
    ├─► "Login" / "Sign in" / "Authenticate"
    │       └─► Load references/login.md
    ├─► "Find posts" / "Research" / "Discover" / "Find people"
    │       └─► Load references/research.md
    ├─► "Comment" / "Post" / "Engage" / "Reply"
    │       └─► Load references/engage.md
    └─► Unclear
            └─► Ask: "Would you like me to research LinkedIn posts/people or engage with existing content?"

Browser MCP Quick Reference

Get Browser Context

tabs_context_mcp (get available tabs)
tabs_create_mcp (create new tab if needed)

Navigation

navigate (url, tabId)

Reading Page

read_page (tabId) - Get page structure
get_page_text (tabId) - Get text content
screenshot - Visual verification

Finding Elements

find (query, tabId) - Natural language element search

Interaction

form_input (ref, value, tabId) - Fill form fields
computer (action: "left_click", coordinate, tabId) - Click elements
computer (action: "scroll", scroll_direction, tabId) - Scroll page
computer (action: "type", text, tabId) - Type text

Important Notes

  1. Authentication: Use login skill or ensure manual login before engaging
  2. Rate Limits: LinkedIn limits connection requests and posting frequency
  3. Professional Tone: LinkedIn rewards professional, insightful content
  4. Connection Limits: Avoid mass connection requests - focus on quality engagement
  5. Content Visibility: LinkedIn algorithm favors comments and engagement within the first hour
  6. User Approval: ALL engagement actions require explicit user confirmation
Weekly Installs
9
GitHub Stars
6
First Seen
Feb 11, 2026
Installed on
opencode9
gemini-cli9
claude-code9
github-copilot9
codex9
kimi-cli9