agentsclimarketplace

Linkedin marketing

Skill ComeOnOliver/skillshub/skills/TerminalSkills/skills/linkedin-marketing

🧠 The right skill, one API call. AI agent skills registry with token-efficient skill resolution. 5,000+ skills from 500+ top repos.

Install
npx -y skills add ComeOnOliver/skillshub --skill linkedin-marketing

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

10.5 KB, ~2.6k tokens by cl100k_base, as published. Nobody here has run it

LinkedIn Marketing

Overview

This skill helps AI agents create high-performing LinkedIn content and integrate with the LinkedIn API. It covers platform-specific content rules, post formatting, carousel creation, API-based publishing, analytics tracking, and growth strategies tailored to LinkedIn's algorithm and audience behavior.

Instructions

Platform Rules & Algorithm

LinkedIn's algorithm in 2025-2026 prioritizes:

  • Dwell time β€” how long people spend reading your post (long-form > short)
  • Early engagement β€” first 60-90 minutes are critical for distribution
  • Comments > reactions β€” comments weight 5-8x more than likes
  • Native content β€” no external links in post body (kills reach 40-50%)
  • Relevance β€” content shown to people interested in the topic, not just connections
  • Consistency β€” 3-5 posts/week beats 1 viral post/month

What kills reach:

  • External links in post body (put in comments instead)
  • Editing post within first hour
  • Engagement pods (LinkedIn detects and suppresses)
  • Hashtag stuffing (3-5 max)
  • Tagging people who don't engage back

Content Formats & Best Practices

Text Posts (Highest organic reach)

[Hook β€” first 2 lines are everything, must stop the scroll]

[Line break after every 1-2 sentences for mobile readability]

[Body β€” story, insight, or framework. 800-1300 characters optimal]

[Each paragraph is 1-3 short sentences max]

[CTA β€” question to drive comments, not "like if you agree"]

[3-5 relevant hashtags at the bottom]

Formatting rules:

  • First 2 lines show before "...see more" β€” make them irresistible
  • Use line breaks aggressively β€” wall of text = scroll past
  • 800-1300 characters is the sweet spot (long enough for dwell time, short enough to finish)
  • Emojis sparingly as bullet markers (πŸ”Ή, β†’, βœ…), not decoration
  • No markdown β€” LinkedIn doesn't render it. Use CAPS or emojis for emphasis
  • End with a question to drive comments (open-ended, not yes/no)

Carousels (PDF documents β€” highest save rate)

  • Upload as PDF (each page is a slide)
  • Slide size: 1080x1350px (4:5 ratio) or 1080x1080px (1:1)
  • 8-12 slides optimal
  • First slide = hook/title (bold, minimal text)
  • Last slide = CTA + profile summary
  • One idea per slide, large text (24-32pt minimum)
  • Brand colors and consistent design
  • File size: under 100MB, max 300 pages

Polls (High engagement, low effort)

  • 2 weeks duration optimal
  • 3-4 options (not 2 β€” too binary)
  • Add context in the post body explaining why you're asking
  • Follow up with results analysis post

Articles & Newsletters

  • Articles: 1000-2000 words, SEO-indexed by Google
  • Newsletters: subscribers get email + in-app notification
  • Good for long-form thought leadership
  • Lower viral potential but higher authority building
  • Include images every 300-400 words

Video

  • Native video (not YouTube links)
  • Vertical (9:16) or square (1:1) β€” NOT widescreen
  • Under 2 minutes for feed (30-90s optimal)
  • Captions mandatory (80%+ watch without sound)
  • Hook in first 3 seconds
  • Max file size: 5GB, max duration: 15 minutes

LinkedIn API Integration

Authentication (OAuth 2.0)

// LinkedIn uses OAuth 2.0 with 3-legged flow
const LINKEDIN_AUTH_URL = 'https://www.linkedin.com/oauth/v2/authorization';
const LINKEDIN_TOKEN_URL = 'https://www.linkedin.com/oauth/v2/accessToken';

// Step 1: Redirect user to authorize
const authUrl = new URL(LINKEDIN_AUTH_URL);
authUrl.searchParams.set('response_type', 'code');
authUrl.searchParams.set('client_id', process.env.LINKEDIN_CLIENT_ID);
authUrl.searchParams.set('redirect_uri', process.env.LINKEDIN_REDIRECT_URI);
authUrl.searchParams.set('scope', 'openid profile email w_member_social');
// Redirect user to authUrl.toString()

// Step 2: Exchange code for token
const tokenResponse = await fetch(LINKEDIN_TOKEN_URL, {
  method: 'POST',
  headers: { 'Content-Type': 'application/x-www-form-urlencoded' },
  body: new URLSearchParams({
    grant_type: 'authorization_code',
    code: authorizationCode,
    client_id: process.env.LINKEDIN_CLIENT_ID,
    client_secret: process.env.LINKEDIN_CLIENT_SECRET,
    redirect_uri: process.env.LINKEDIN_REDIRECT_URI,
  }),
});
const { access_token, expires_in } = await tokenResponse.json();
// Token valid for 60 days, refresh before expiry

Post Content via API

// Get user profile URN
const profileRes = await fetch('https://api.linkedin.com/v2/userinfo', {
  headers: { Authorization: `Bearer ${access_token}` },
});
const { sub: personUrn } = await profileRes.json();

// Create a text post
const postRes = await fetch('https://api.linkedin.com/v2/posts', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${access_token}`,
    'Content-Type': 'application/json',
    'X-Restli-Protocol-Version': '2.0.0',
    'LinkedIn-Version': '202401',
  },
  body: JSON.stringify({
    author: `urn:li:person:${personUrn}`,
    commentary: 'Your post text here...',
    visibility: 'PUBLIC',
    distribution: {
      feedDistribution: 'MAIN_FEED',
      targetEntities: [],
      thirdPartyDistributionChannels: [],
    },
    lifecycleState: 'PUBLISHED',
  }),
});

// Create a post with image
// Step 1: Initialize upload
const uploadRes = await fetch('https://api.linkedin.com/v2/images?action=initializeUpload', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${access_token}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    initializeUploadRequest: {
      owner: `urn:li:person:${personUrn}`,
    },
  }),
});
const { value: { uploadUrl, image: imageUrn } } = await uploadRes.json();

// Step 2: Upload image binary
await fetch(uploadUrl, {
  method: 'PUT',
  headers: { 'Content-Type': 'image/png' },
  body: imageBuffer,
});

// Step 3: Create post with image
await fetch('https://api.linkedin.com/v2/posts', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${access_token}`,
    'Content-Type': 'application/json',
    'LinkedIn-Version': '202401',
  },
  body: JSON.stringify({
    author: `urn:li:person:${personUrn}`,
    commentary: 'Post with image...',
    visibility: 'PUBLIC',
    distribution: {
      feedDistribution: 'MAIN_FEED',
    },
    content: {
      media: {
        id: imageUrn,
        title: 'Image title',
      },
    },
    lifecycleState: 'PUBLISHED',
  }),
});

Company Page Posting

// Post as company page (requires r_organization_admin + w_organization_social)
await fetch('https://api.linkedin.com/v2/posts', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${access_token}`,
    'Content-Type': 'application/json',
    'LinkedIn-Version': '202401',
  },
  body: JSON.stringify({
    author: 'urn:li:organization:12345678',
    commentary: 'Company update...',
    visibility: 'PUBLIC',
    distribution: { feedDistribution: 'MAIN_FEED' },
    lifecycleState: 'PUBLISHED',
  }),
});

Analytics

// Get post analytics (requires r_organization_social scope for org posts)
const analyticsRes = await fetch(
  `https://api.linkedin.com/v2/organizationalEntityShareStatistics?q=organizationalEntity&organizationalEntity=urn:li:organization:12345678&timeIntervals.timeGranularityType=DAY&timeIntervals.timeRange.start=1704067200000&timeIntervals.timeRange.end=1706745600000`,
  { headers: { Authorization: `Bearer ${access_token}` } }
);

// Personal post metrics via ugcPosts
const ugcRes = await fetch(
  `https://api.linkedin.com/v2/socialMetadata/${postUrn}`,
  { headers: { Authorization: `Bearer ${access_token}` } }
);

Growth Strategy

Posting schedule: 3-5x/week (Tue-Thu peak engagement, 8-10 AM local time)

Content mix:

  • 40% educational (frameworks, how-tos, lessons learned)
  • 25% stories (personal experiences, behind-the-scenes)
  • 20% opinion/contrarian takes (drives comments)
  • 10% engagement posts (polls, questions)
  • 5% promotional (product, hiring, announcements)

Comment strategy:

  • Reply to every comment on your posts within 2 hours
  • Leave 5-10 thoughtful comments/day on posts in your niche
  • Comments on big accounts get you visibility to their audience
  • Ask follow-up questions in comments to extend threads

Examples

Example 1: Write a LinkedIn post announcing a product launch

User prompt: "Write a LinkedIn post for me. I'm the CTO of Meridian Analytics and we just launched a real-time data pipeline product called StreamFlow that processes 2 million events per second. Target audience is engineering leaders at mid-size SaaS companies."

The agent will draft a LinkedIn text post optimized for the algorithm. The first two lines hook with: "We spent 14 months building something we couldn't find anywhere else.\n\nToday we're launching StreamFlow β€” a real-time data pipeline that handles 2M events/sec with sub-10ms latency." The body tells the story of the problem (batch processing delays costing customers $40K/month in stale dashboards), the solution, and early results from beta customers. It ends with an open-ended question: "What's the biggest bottleneck in your data pipeline today?" followed by 4 hashtags: #DataEngineering #RealTimeAnalytics #SaaS #ProductLaunch. The link to the product page goes in a separate first comment, not in the post body.

Example 2: Automate posting to a LinkedIn company page via API

User prompt: "Build a TypeScript function that posts a weekly engineering blog summary to our LinkedIn company page (Meridian Analytics, org ID 91847362) every Monday at 9 AM EST."

The agent will create a TypeScript module that authenticates with LinkedIn OAuth 2.0 using LINKEDIN_CLIENT_ID and LINKEDIN_CLIENT_SECRET from environment variables, obtains an access token with w_organization_social scope, then calls POST /v2/posts with author: "urn:li:organization:91847362", the formatted blog summary as commentary, visibility: "PUBLIC", and lifecycleState: "PUBLISHED". The function is wrapped in a cron scheduler using node-cron with the pattern 0 9 * * 1 (Mondays at 9 AM). It includes token refresh logic to handle the 60-day expiry and logs each post's success or failure to a linkedin-posts.log file.

Guidelines

  • Put links in the FIRST COMMENT, never in the post body
  • Write for mobile β€” 60%+ of LinkedIn users are on phones
  • Don't edit posts in the first hour β€” it resets distribution
  • Use "I" not "we" for personal posts β€” LinkedIn is personal-first
  • Tag max 3-5 people per post, only if they'll actually engage
  • Post consistently at the same times β€” your audience learns when to expect you
  • Repurpose high-performing posts after 3-4 months β€” new followers haven't seen them
  • Save carousels as PDF at 1080x1350px for optimal mobile display

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.