SkillsApr 13, 2026·1 min read

Multi-Platform Social Media Growth Skill for Developer Communities

Production-tested Claude Code skill for automating authentic developer engagement across 10+ platforms (Twitter, Bluesky, Dev.to, HN, Reddit, GitHub, Product Hunt, Hashnode, Medium, Quora). Includes rotation queue, per-platform cooldowns, action diversity rules, Chrome MCP toolchain with JS injection patterns, anti-ban safety, and continuous optimization framework. Battle-tested across 30+ cycles.

HE
henuwangkai · Community
Quick Use

Use it first, then decide how deep to go

This block should tell both the user and the agent what to copy, install, and apply first.

Intro

This skill automates continuous engagement across developer social platforms using a rotating queue system. It's battle-tested across 30+ cycles and covers persona management, content strategy, platform-specific tooling, rate limiting, and anti-ban safety.

Use case: Indie hackers, developer advocates, or AI tool builders who want to grow presence across multiple dev communities without manual effort.


1. Platform Queue & Rotation

Supported Platforms (10 active)

Dev.to → Quora → Twitter/X → Bluesky → Hacker News → Reddit → GitHub → Product Hunt → Hashnode → Medium → (loop)

Rotation Rules

  1. Each trigger executes 3 platforms, then stops
  2. Next trigger resumes from breakpoint (e.g., stopped at Bluesky → next starts at HN)
  3. Breakpoint can cross cycle boundaries (Medium done → next cycle starts at Dev.to)
  4. Platforms execute back-to-back within a trigger — no waiting between them
  5. Each cycle uses a fresh content angle (never repeat the previous cycle's theme)

2. Frequency & Cooldown

Cron Trigger

  • Every 30 minutes (*/30 * * * *)
  • Cron is a fallback — if already executing, skip

Per-Platform Posting Cooldowns

Platform Post/Original Cooldown Daily Post Cap Engagement Cooldown
Twitter/X ≥2 hours ~6 tweets None
Bluesky ≥1.5 hours ~8 posts None
Dev.to ≥3 hours ~4 comments None
Quora ≥2 hours ~6 answers None
Medium ≥4 hours ~3 responses None
HN/Reddit/GitHub/PH/Hashnode ≥1 hour ~10 each None

During cooldown: Only perform engagement actions (likes/upvotes/stars/follows), no new posts/comments.


3. Action Diversity Rules

Iron rule: Each platform requires ≥3 different action types per visit (not 3x the same type)

Platform Action Menus

Platform Available Actions
Twitter/X Like, Reply, Follow, Original Post, Quote Tweet, Retweet, Bookmark, Join Community, DM
Bluesky Like, Reply, Repost, Quote post, Follow, Original post
Dev.to Heart, Unicorn, Bookmark, Follow author, Comment, Reply to comments
Quora Upvote, Answer, Comment, Follow user, Follow topic
Hacker News Upvote, Comment
Reddit Upvote, Comment, Follow, Join subreddit
GitHub Star repo, Follow user, Issue comment, Discussion comment
Product Hunt Upvote, Comment, Follow maker
Hashnode Upvote/Like, Comment, Follow user, Bookmark
Medium Clap, Highlight, Respond, Follow author

Minimum Per Visit

  • 2-3 likes/upvotes (baseline)
  • At least 1 "light social" action: Follow / Bookmark / Repost / Star
  • At least 1 "content" action: Reply / Original post / Comment / Answer
  • Every 2-3 cycles: 1 "deep" action: DM / long article / Twitter Thread / join community

4. Content Strategy

Content Pillars

Type Share Examples
AI Industry Insights 40% Unique takes on AI trends
Tool Recommendations & Tips 30% Claude Code tips, MCP configs, prompt engineering
Product/Asset Sharing 20% Natural mention of your product in context
Community Engagement 10% Questions, discussions, responding to others

Link Mention Rules

  • ~50% of original content naturally mentions your product URL
  • Mentions must feel organic: sharing discoveries/recommendations/experiences, never ad-like
  • Other 50% is pure value content with no links (maintains credibility)
  • Max 30% of total activity should contain links

Content Angle Rotation (change each cycle)

Example angles used in production:

  • Trust vs verify AI output
  • AI pair programming: lead vs follow
  • Shipping real AI tools / indie hacker speed
  • Dev tools product-market fit
  • Open source as distribution strategy
  • AI testing & quality gap
  • DX polish as competitive advantage
  • Technical debt in AI-generated code

Platform Adaptation

  • Twitter: 280 char limit, punchy, 1-3 hashtags
  • Bluesky: 300 char limit
  • Dev.to/Hashnode: Long comments/articles, technical depth
  • Quora: Answer format, personal experience + data
  • Medium: Response format, insightful
  • HN: Minimal technical style, no emoji or hashtags
  • Reddit: Match subreddit culture, avoid over-promotion
  • GitHub: Technical comments, code-related
  • Product Hunt: Product discussion perspective

5. Chrome MCP Toolchain

Only use mcp__beta__* tools (not mcp__chrome__*)

Core Workflow

take_snapshot → save to /tmp/xxx.txt → Grep for UIDs → interact

Key Technical Patterns

Universal Snapshot Flow

  1. mcp__beta__take_snapshot → capture page DOM
  2. Save to /tmp/{platform}-c{cycle}.txt
  3. Grep search for target element UIDs
  4. Execute click / fill / type_text using UIDs

Twitter/X Posting

  • Navigate to /compose/post for clean composer
  • Clear existing text: Meta+A + Backspace (via press_key)
  • Check length: evaluate_script reading textbox textContent.length

Dev.to Comment Injection

() => {
  const ta = document.getElementById('text-area');
  if (!ta) return 'not found';
  const nativeSet = Object.getOwnPropertyDescriptor(
    Object.getPrototypeOf(ta), 'value'
  ).set;
  nativeSet.call(ta, "YOUR COMMENT TEXT");
  ta.dispatchEvent(new Event('input', {bubbles: true}));
  ta.dispatchEvent(new Event('change', {bubbles: true}));
  return 'done';
}

Medium Contenteditable Focus

() => {
  const ces = document.querySelectorAll('[contenteditable="true"]');
  for (const ce of ces) {
    if (ce.closest('[role="textbox"]') || 
        ce.textContent.includes('What are your thoughts')) {
      ce.focus();
      const range = document.createRange();
      range.selectNodeContents(ce);
      range.collapse(false);
      const sel = window.getSelection();
      sel.removeAllRanges();
      sel.addRange(range);
      return 'focused';
    }
  }
  if (ces.length > 0) { ces[0].focus(); return 'focused first'; }
  return 'not found';
}

Quora Editor Focus

() => {
  const editor = document.querySelector('[contenteditable="true"].doc');
  if (!editor) return 'not found';
  editor.focus();
  const range = document.createRange();
  range.selectNodeContents(editor);
  range.collapse(false);
  const sel = window.getSelection();
  sel.removeAllRanges();
  sel.addRange(range);
  return 'focused';
}

6. Safety & Anti-Ban Rules

Rate Limiting

  • Don't rapid-fire likes (spread evenly)
  • Never post identical text across platforms
  • Link content <30%, 70% pure engagement/value
  • If rate limited or warned → stop that platform for 24 hours
  • Human-like timing: don't post at exact intervals

Platform Link Restrictions (tested)

Platform Link Status
Twitter/Bluesky/Dev.to/HN/Reddit/Quora/Hashnode/Medium URLs allowed
Product Hunt comments URLs blocked (triggers "invalid content")
IndieHackers New accounts can't post URLs
Lobsters Invite-only registration

Account Management

  • Prefer Google OAuth for new signups
  • Consistent username/display name across platforms
  • Consistent bio mentioning your product
  • Only ask for human help on CAPTCHA/SMS verification

7. Continuous Optimization

  1. Review every 3-5 cycles: Which platforms/angles got engagement?
  2. Content evolution: Rotate topics based on platform trending, never repeat
  3. Platform deep-dive: Invest more in high-conversion platforms (Dev.to long posts, Quora long answers, HN technical discussions)
  4. Traffic tracking: Periodically check analytics and per-platform engagement data
  5. Skill iteration: Update this skill when discovering new effective strategies
  6. Deep actions: Execute one deep action every 2-3 cycles (long article / Thread / DM / join community)

8. Localized UI Label Mapping

If your browser is in a non-English locale, common button labels:

Local Label English Platform
发帖 Post General
回复 Reply General
喜欢 Like Bluesky
创建 Create Threads
发布 Publish General
撰写新帖文 Compose new post Bluesky
发布帖文 Publish post Bluesky
星星 Star GitHub

Credits

Production-tested across 30+ complete rotation cycles covering 10 platforms with 3+ action types each. Developed for promoting developer tools in AI/productivity niche.

Discussion

Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.