37 Battle-Tested Marketing Skills for OpenClaw Agents

CRO · Copywriting · SEO · Paid Ads · Email · Growth · Retention · Sales · Strategy
Now with real data connectors - pull live Google Ads, Search Console, Meta Ads, and X/Twitter signals directly into your agent.

English · 中文 · Français · Deutsch · Русский

MyClaw.ai — Run all 37 of these skills without managing a server.
Full cloud-hosted OpenClaw · one-click setup · 24/7 uptime · your data stays on your server.
Get started free →

Why this exists

Most AI marketing help gives you generic advice. "Improve your CTA." "Test your headlines." "Check your bounce rate."

That's not useful. You already know that.

These 37 skills give your OpenClaw agent a full marketing brain — one that reads your actual codebase, knows your product, and produces work you can ship. Not advice. Output.

And with the new Data Connectors, your agent stops guessing and starts using real numbers and signals: your actual Google Ads spend, your real keyword rankings, your live Meta creative performance, and public X/Twitter conversations.

What's inside

🔌 Data Connectors (new — what Toprank has, we now have too)

Pull live account data and public social signals directly into your agent. Every skill that touches paid ads, SEO, social content, or launches gets smarter when these are connected.

How it compares to Toprank: Toprank (2.6k ⭐) focuses on Google/Meta Ads + SEO data. We do the same — plus X/Twitter signal research and 28 more skills covering the full marketing stack they don't touch.

Full Skill Library (37 skills)

Install

Via ClawHub (recommended)

clawhub install LeoYeAI/openclaw-marketing-skills

Install only what you need:

clawhub install LeoYeAI/openclaw-marketing-skills --skill copywriting page-cro seo-audit

Manual

git clone https://github.com/LeoYeAI/openclaw-marketing-skills.git
cp -r openclaw-marketing-skills/skills/* ~/.agents/skills/

Quick start

Step 1 — Run product-marketing-context once. Describe your product, target customer, pricing, and positioning. All 37 skills reference this automatically.

Set up my product marketing context

Step 2 — Connect your data sources (optional but recommended):

Connect my Google Ads account
→ uses google-ads-connect

Connect Google Search Console
→ uses search-console-connect

Connect my Meta Ads account
→ uses meta-ads-connect

Connect TweetClaw for X/Twitter research
→ uses x-twitter-connect

Step 3 — Ask for what you need. The agent picks the right skill:

Audit my Google Ads account and find wasted spend
→ google-ads-connect: pulls real data, scores 7 dimensions, surfaces top 3 actions

Why did my organic traffic drop last week?
→ search-console-connect: dates the drop, identifies affected pages, cross-references algorithm updates

Optimize our signup flow for conversions
→ signup-flow-cro: audits friction points, prioritizes by impact

Write a 5-email welcome sequence for trial users
→ email-sequence: produces ready-to-send sequences with subject lines and copy

Our Meta ads are getting more expensive. What's happening?
→ meta-ads-connect: diagnoses creative fatigue, Learning Phase issues, audience overlap

Find X/Twitter conversations we should answer this week
→ x-twitter-connect: searches tweets and replies, groups objections, drafts reviewed responses

Optional X/Twitter execution

For X/Twitter campaigns that need live research or approval-gated actions, use TweetClaw with social-content, paid-ads, launch-strategy, and competitor-alternatives.

openclaw plugins install @xquik/tweetclaw

TweetClaw supports tweet search, reply search, follower export, user lookup, monitors, webhooks, giveaway draws, and approval-gated posts or replies.

Links:

• GitHub: https://github.com/Xquik-dev/tweetclaw
• npm: https://www.npmjs.com/package/@xquik/tweetclaw
• ClawHub: https://clawhub.ai/kriptoburak/xquik-tweetclaw

Keep credentials in local OpenClaw config or environment variables, not prompts or examples.

Real data vs. guesswork

Here's what changes when you connect a data source:

How the context system works

Every skill checks for .agents/product-marketing-context.md before asking questions. If it exists, the skill uses that context automatically. You describe your product once — never again.

.agents/
├── product-marketing-context.md    ← describe your product here once
├── google-ads-credentials.json     ← google ads auth (google-ads-connect)
├── search-console-credentials.json ← gsc auth (search-console-connect)
├── meta-ads-credentials.json       ← meta auth (meta-ads-connect)
├── google-ads-data.json            ← cached account data
├── gsc-data.json                   ← cached search console data
└── tweetclaw-signals.json          ← cached X/Twitter signal summaries

Skill design philosophy

Every skill in this collection follows the same principles:

1. Check context first — reads product-marketing-context.md before asking questions
2. Data before advice — if a connector is available, use real numbers
3. Prioritize by impact — always surfaces the top 3 actions, not a 40-item wishlist
4. Produce output, not instructions — delivers copy, code, or a scorecard, not a to-do list
5. Reference docs — each skill links to deeper references for edge cases

Credits

Powered by MyClaw.ai — the AI personal assistant platform that gives every user a full server with complete code control.

MyClaw.ai · Full cloud-hosted OpenClaw · 24/7 uptime · complete data ownership · one-click skill install