
Marketing Skills for AI Agents
github.com/coreyhaines31/marketingskills- Category
- AI Agents
- Rank
- No. 173Tools index
- Pricing
- Open Source
- Type
- TOOL
- Use case
- Business & Commerce
- Interfaces
- Agent Skill / Plugin
- Builder
- @coreyhaines31
- GitHub
- 51.4k stars
- Latest release
- v2.11.1
- Date
About
A collection of specialized marketing skills for AI coding agents, covering CRO, copywriting, SEO, analytics, and growth engineering. Built for technical marketers and founders who want AI agents to help with marketing tasks using structured markdown skills that provide frameworks and best practices.
What it does
It gives an agent a shared product brief, then routes marketing requests into focused playbooks. Those playbooks reference related disciplines, so research can inform copy, conversion work can feed experiments, and sales tasks can share context. Optional command-line adapters connect selected workflows to external marketing services.
Why it's ranked here
The strongest idea is coordination, not sheer topic count. A common product context reduces disconnected advice, while explicit links between disciplines encourage coherent campaigns. The included service adapters also move some work beyond recommendations into execution. However, the supplied material demonstrates breadth more clearly than the depth or validation of each playbook.
What's good
The dependency model teaches agents to establish audience, positioning, and product facts before producing tactics. Cross-references encode useful handoffs among research, copy, conversion work, experiments, SEO, and sales. The command-line adapters use consistent JSON output, validate many required inputs, read credentials from environment variables, and commonly provide dry-run previews with secrets masked.
Tradeoffs
Most value still depends on the host agent interpreting markdown instructions well. External actions require separate vendor accounts and credentials. The adapters repeat argument parsing and request handling instead of sharing a common runtime. Several wrappers return parsed response bodies without treating unsuccessful HTTP status codes as failures, so automation must inspect results carefully.
How to use it well
It suits technical marketers and founders already working inside a compatible coding agent. Start by recording product, audience, and positioning context, then invoke one focused playbook and follow its related handoffs. Use dry runs before external mutations. It does not replace vendor access, campaign approval, customer evidence, or independent measurement of results.
Technical notes+
README.md defines the shared-context dependency model and lists compatibility with Claude Code, OpenAI Codex, Cursor, Windsurf, and Agent Skills spec hosts. The supplied adapters under tools/clis/ are executable Node.js scripts with hand-written argument parsing, environment-variable authentication, fetch-based HTTP requests, and JSON stdout. tools/clis/dub.js, tools/clis/ga4.js, tools/clis/kit.js, tools/clis/clay.js, tools/clis/brevo.js, tools/clis/close.js, tools/clis/demio.js, tools/clis/airops.js, and tools/clis/apollo.js expose mutating operations. tools/clis/exa.js, tools/clis/g2.js, tools/clis/pendo.js, and tools/clis/ahrefs.js emphasize search, reporting, or retrieval. Dry-run branches mask credentials, but the wrappers generally parse response text without checking res.ok.
Observed
- Primary formats
- Markdown agent skills with JavaScript command-line adapters.
- Agent interface
- Compatible with Claude Code, OpenAI Codex, Cursor, Windsurf, and Agent Skills spec hosts.
- Execution interface
- Executable Node.js command-line scripts call external vendor APIs and print JSON.
- Authentication
- Vendor credentials are supplied through environment variables.
- Safety surface
- Many adapters provide a dry-run option that previews requests and masks credentials.
- Architecture
- A shared product context feeds specialized skills, which explicitly cross-reference related marketing disciplines.
Read from README.md, tools/clis/g2.js, tools/clis/dub.js, tools/clis/exa.js, tools/clis/ga4.js, tools/clis/kit.js, tools/clis/clay.js, tools/clis/brevo.js, tools/clis/close.js, tools/clis/demio.js, tools/clis/pendo.js, tools/clis/ahrefs.js, tools/clis/airops.js, tools/clis/apollo.js.
What it can do
Optimize conversion rates for websites and landing pages
Website or landing page data and conversion metrics → CRO recommendations and optimization strategies
Generate marketing copy and content
Product information and target audience details → Marketing copy, headlines, and content variations
Analyze and optimize SEO performance
Website content and SEO requirements → SEO optimization recommendations and keyword strategies
Process and analyze marketing analytics data
Marketing performance data and metrics → Analytics insights and performance reports
Execute growth engineering strategies
Product and user data → Growth tactics and engineering solutions for user acquisition
Create structured marketing workflows for AI agents
Marketing requirements and business context → Markdown-formatted marketing skills and frameworks
Tags
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.