AI Marketing Skills
ericosiu/ai-marketing-skillsReviewed by AI
Checked by RepoPicks Radar on
What is AI Marketing Skills?
This repository collects around 30 ready-made "skills" for Claude Code and similar AI assistants, written by a marketing agency. Each folder holds a SKILL.md plus Python scripts for one job: growth experiments, turning website visitors into sales leads, scoring content, cold outbound email, SEO briefs, YouTube thumbnails, short-form video editing with FFmpeg, finance reports. You copy the folder you want into your assistant's skills directory, install its Python packages and fill in a .env file with keys for the outside services it uses.
How easy is it to install AI Marketing Skills?
C · install 1 or 2 tools · Windows
- Windows C
- macOS C
- Linux C
The same work on all three systems, and it is a C: you clone the repository, copy a whole skill folder into .claude/skills, then per the Quick Start "pip install -r requirements.txt", copy .env.example to .env and add keys before running commands such as "python experiment-engine.py create --hypothesis ...". Some video skills also need FFmpeg installed.
AI Marketing Skills on GitHub
- Stars
- 3,803on Nov 10, 2026
- Downloads
- —all releases
- Growth
- +23stars in 6 days
Stars over time
Facts
- Runs on
- Windows, macOS, Linux
- You need to install
- Python, FFmpeg
- To build from source
- FFmpeg
- Disk space
- not determined
- Accounts or keys
- Google Gemini (video understanding skills), Google Search Console (SEO skills), Gong (revenue intelligence skills), RB2B and Instantly (outbound skills), Metricool (short-form delivery skill)
- Cost
- Free, API costs extra · The skills are MIT licensed and free. They only run inside an AI assistant such as Claude Code (normally a paid plan), and several of them call paid services (for example Gong, RB2B, Instantly, Metricool, Gemini).
- License
- MIT (Permissive)
- Can replace
- Semrush, PhantomBuster, Jasper, SEO
- Last change on GitHub
- Nov 10, 2026
AI Marketing Skills risk signals (5)
- Attentionautomated cold outreach and visitor tracking
The Outbound Engine is described as "ICP definition to emails in inbox — fully automated", Sales Pipeline promises to "Turn anonymous website visitors into qualified pipeline" with an "RB2B Router" that does "intent scoring, seniority-based company dedup, and agency classification before routing to outbound sequences", and "Deal Resurrector" includes "tracking departed contacts to their new companies". Sending bulk cold email and identifying people who browsed your site are limited by spam and data-protection laws in many countries, including the GDPR in Europe and Brazil's LGPD.
AI review
- Attentionskills can publish and send on your behalf
Several skills hand an AI agent the keys to live accounts and channels: "Shortform Production ... deliver approved clips through the API | Metricool Delivery", "Deck Generator ... Google Slides Builder", and the growth engine runs experiments that "run, measure, and optimize themselves". A mistake is published or emailed to real people, not kept in a draft folder.
AI review
- Inforepository promotes the vendor services
The README is wrapped in the agency's funnel: links to singlebrain.com and singlegrain.com carry "utm_source=github&utm_medium=skill_repo&utm_campaign=ai_marketing_skills", plus "Want these built and managed for you?" and a newsletter sign-up. Claims like "battle-tested on real pipelines generating millions in revenue" come from the vendor.
AI review
- Infotelemetry described as opt in
README Telemetry section: "Local logging always — see your own usage stats in ~/.ai-marketing-skills/analytics/", "Remote reporting optional — only if you explicitly opt in on first run", collecting "skill name, duration, success/fail, version, OS". It also runs version checks (telemetry/version_check.py).
AI review
- Infono single install each skill differs
The repository has about 30 top-level skill folders, no releases and no root install: "Each skill category has its own README with setup instructions", each with its own requirements.txt and .env.example, so what you must set up changes from skill to skill.
AI review
Evidence is shown in its original language.
What Radar could not check about AI Marketing Skills
- Whether each script still works against the outside services it calls (Gong, RB2B, Instantly, Metricool, Search Console, Gemini), since none of them is pinned or tested here
- What a full run costs in AI tokens and in subscriptions to those services
- Whether the statistical claims (bootstrap confidence intervals, Mann-Whitney U tests) are implemented correctly
- Whether the PII sanitizer really catches your customers' data before it leaves your machine
- How much disk space and how much of your own data each skill needs, which the checks could not measure