Sales & marketingkitMIT

AgentKits Marketing

A marketing command suite for your agent: campaigns, SEO, CRO, competitor research, and reporting.

773 files2.1 MB572 stars on GitHub41 from Neura Market

Original work

Created by
aitytech
Published on
GitHub
License
MIT

Mirrored on Neura Market with prose references to the original marketplace updated to our name. Install commands and other URLs inside the files are left untouched so they keep working. All credit for the kit belongs to its author.

What's inside

773 files in this kit.

  • .claude-plugin/marketplace.json

    File
  • .claude-plugin/plugin.json

    File
  • .claude/.env.example

    File
  • .claude/.gitignore

    File
  • .claude/.mcp.json.example

    File
  • .claude/agents

    File
  • .claude/commands

    File
  • .claude/components/date-helpers.md

    Documentation
  • .claude/components/interactive-questions.md

    Documentation
  • .claude/metadata.json

    File
  • .claude/settings.json

    File
  • .claude/settings.local.json

    File
  • .claude/skills

    File
  • .claude/statusline.js

    File
  • .claude/statusline.ps1

    File
  • .claude/statusline.sh

    File
  • .claude/workflows/crm-workflow.md

    Documentation
  • .claude/workflows/data-reliability-rules.md

    Documentation
  • .claude/workflows/documentation-management.md

    Documentation
  • .claude/workflows/marketing-rules.md

    Documentation
  • .claude/workflows/orchestration-protocol.md

    Documentation
  • .claude/workflows/primary-workflow.md

    Documentation
  • .claude/workflows/sales-workflow.md

    Documentation
  • .commitlintrc.json

    File
  • .env.example

    File
  • .github/CODEOWNERS

    File
  • .github/FUNDING.yml

    File
  • .github/ISSUE_TEMPLATE/bug_report.yml

    File
  • .github/ISSUE_TEMPLATE/config.yml

    File
  • .github/ISSUE_TEMPLATE/feature_request.yml

    File
  • .github/ISSUE_TEMPLATE/question.yml

    File
  • .github/PULL_REQUEST_TEMPLATE.md

    Documentation
  • .github/SECURITY.md

    Documentation
  • .github/dependabot.yml

    File
  • .github/labeler-config.yml

    File
  • .github/labels.yml

    File
  • .github/mlc_config.json

    File
  • .github/release-drafter.yml

    File
  • .github/workflows/ci.yml

    File
  • .github/workflows/labeler.yml

    File
  • .github/workflows/release-drafter.yml

    File
  • .github/workflows/release.yml

    File
  • .github/workflows/stale.yml

    File
  • .github/workflows/welcome.yml

    File
  • .gitignore

    File
  • .husky/commit-msg

    File
  • .releaserc.json

    File
  • .repomixignore

    File
  • .vscode/settings.json

    File
  • CHANGELOG.md

    Documentation
  • CLAUDE.md

    Documentation
  • CODEOWNERS

    File
  • CONTRIBUTING.md

    Documentation
  • GEMINI.md

    Documentation
  • LICENSE

    File
  • README.ar.md

    Documentation
  • README.de.md

    Documentation
  • README.es.md

    Documentation
  • README.fr.md

    Documentation
  • README.ja.md

    Documentation

…and 713 more files in the download.

About this kit

AgentKits Marketing Toolkit Reference Guide

Overview

The AgentKits Marketing Toolkit is an enterprise-grade AI marketing automation toolkit designed for AI assistants such as Claude Code, Cursor, GitHub Copilot, Windsurf, and Cline. It provides specialized agents, skills, commands, and workflows for SaaS founders, marketers, and growth teams. The toolkit automates campaign planning, content creation, SEO, CRO, email sequences, and analytics. It is open source under the MIT license and supports self-hosted marketplace via marketplace.json.

When to Use

Use this toolkit when you need to automate marketing tasks using an AI assistant that supports agents and skills. It is intended for SaaS founders, marketers, and growth team members looking for production-ready marketing agents and workflows. Install it as a comprehensive marketing plugin for Claude Code, Cursor, GitHub Copilot, Windsurf, or Cline. Use it to run marketing campaigns, create content, optimize conversions, or perform SEO at scale. Learn AI-powered marketing through interactive training lessons. Reference a library of marketing best practices, frameworks, and templates.

Capabilities

Specialized Marketing Agents

The toolkit provides 18 specialized marketing agents:

  • attraction-specialist
  • lead-qualifier
  • email-wizard
  • sales-enabler
  • continuity-specialist
  • upsell-maximizer
  • copywriter
  • conversion-optimizer
  • seo-specialist
  • brand-voice-guardian
  • researcher
  • brainstormer
  • planner
  • project-manager
  • docs-manager
  • solopreneur
  • startup-founder

Slash Commands

There are 93 slash commands organized by category: campaign, content, SEO, CRO, growth, email, analytics, sales, research, marketing, and testing. Commands are used directly in the AI assistant chat. Examples include:

  • /campaign:plan
  • /content:blog
  • /cro:page
  • /seo:keywords
  • /growth:launch
  • /pricing:strategy
  • /test:ab-setup

Marketing Skills

The toolkit includes 28 marketing skills:

  • marketing-psychology (70+ mental models)
  • marketing-ideas (140+ SaaS strategies)
  • page-cro
  • form-cro
  • popup-cro
  • signup-flow-cro
  • onboarding-cro
  • paywall-upgrade-cro
  • ab-test-setup
  • copywriting
  • copy-editing
  • email-sequence
  • seo-mastery
  • programmatic-seo
  • schema-markup
  • competitor-alternatives
  • launch-strategy
  • pricing-strategy
  • referral-program
  • free-tool-strategy

Core Marketing Workflows

Three core workflows are provided:

  • primary-workflow for campaign lifecycle
  • sales-workflow for lead to customer
  • crm-workflow for contact lifecycle

Interactive Training

The toolkit includes 22 interactive training lessons totaling 5-6 hours. Training is available in 10 languages: English, Simplified Chinese, Japanese, Korean, Spanish, German, French, Portuguese, Vietnamese, Russian, and Arabic. The training is structured as a practice project where the user acts as a Marketing Strategist at the Markit agency, with the client being AgentKits.

Training modules:

  • Module 0: Getting Started (30 minutes)
  • Module 1: Core Concepts (90 minutes)
  • Module 2: Advanced Applications (120 minutes)
  • Module 3: CRO and Conversion (60 minutes)
  • Bonus content: pattern library, 10x Marketer Framework

MCP Integrations

The toolkit supports MCP integrations with real data services:

  • SensorTower
  • Google Search Console
  • Google Analytics
  • Semrush
  • DataForSEO
  • Meta Ads
  • HubSpot

Learning Paths

Four learning paths are available:

  • Quick Start (30 minutes)
  • Full Training (5-6 hours)
  • Skill-Specific (15-30 minutes each)
  • CRO Mastery (60 minutes)

Installation Procedures

Installation via Claude Code Plugin Marketplace (Recommended for Claude Code)

This method is recommended for Claude Code users. Follow these steps in order:

  1. Add the marketplace: /plugin marketplace add aitytech/agentkits-marketing

  2. Install the full suite: /plugin install agentkits-marketing@agentkits-marketing

  3. Optionally install individual components:

    • Skills only: /plugin install agentkits-marketing-skills@agentkits-marketing
    • Agents only: /plugin install agentkits-marketing-agents@agentkits-marketing
    • Commands only: /plugin install agentkits-marketing-commands@agentkits-marketing
  4. Restart Claude Code after installation.

Installation via npx (All Platforms)

This method works for all supported platforms. Follow these steps:

  1. Run: npx @aitytech/agentkits-marketing install

  2. Optionally specify the platform:

    • npx @aitytech/agentkits-marketing install --platform claude (Claude Code)
    • npx @aitytech/agentkits-marketing install --platform cursor (Cursor IDE)
    • npx @aitytech/agentkits-marketing install --platform windsurf (Windsurf)
    • npx @aitytech/agentkits-marketing install --platform cline (Cline)
    • npx @aitytech/agentkits-marketing install --platform copilot (GitHub Copilot)
    • npx @aitytech/agentkits-marketing install --platform all (all platforms)
  3. Use other CLI commands as needed:

    • npx @aitytech/agentkits-marketing --help (show all commands)
    • npx @aitytech/agentkits-marketing list-ides (list supported IDEs)
    • npx @aitytech/agentkits-marketing list-modules (list available modules)
    • npx @aitytech/agentkits-marketing update (update existing installation)

Installation via Clone and Use

Follow these steps:

  1. Clone the repository: git clone https://github.com/aitytech/agentkits-marketing.git

  2. Change directory: cd agentkits-marketing

  3. Run: claude

Manual Installation

Follow these steps:

  1. Clone the repo: git clone https://github.com/aitytech/agentkits-marketing.git

  2. Copy agents: cp agentkits-marketing/.claude/agents/*.md ~/.claude/agents/

  3. Copy commands: cp -r agentkits-marketing/.claude/commands/* ~/.claude/commands/

  4. Copy skills: cp -r agentkits-marketing/.claude/skills/* ~/.claude/skills/

  5. Copy workflows: cp -r agentkits-marketing/.claude/workflows/* ~/.claude/workflows/

Quick Start Procedures

Quick Start - Campaign Launch

  1. Research and plan:

    • /research:market "SaaS productivity tools"
    • /competitor:deep "competitor.com"
    • /campaign:plan "Q1 Product Launch"
  2. Generate content:

    • /content:landing "new feature" "target audience"
    • /content:email "product launch" "trial users"
    • /content:blog "feature announcement" "primary keyword"
  3. Optimize:

    • /cro:page "landing page for conversion"
    • /seo:optimize "content.md" "target keyword"

Quick Start - Content Creation

  1. /content:good "Blog post about AI marketing"
  2. /content:editing "polish this draft"
  3. /seo:keywords "ai marketing automation"

Quick Start - Conversion Optimization

  1. /cro:page "homepage conversion audit"
  2. /cro:form "lead capture optimization"
  3. /cro:signup "registration flow"
  4. /test:ab-setup "headline variations"

Quick Start - Growth and Strategy

  1. /marketing:ideas "SaaS product"
  2. /marketing:psychology "pricing objections"
  3. /growth:launch "Product Hunt strategy"
  4. /pricing:strategy "tier structure"

Training Procedures

Training - Start in Your Language

Use the appropriate command prefix for your language:

  • English: /training:start-0-0
  • Chinese: /training-zh:start-0-0
  • Japanese: /training-ja:start-0-0
  • Korean: /training-ko:start-0-0
  • Spanish: /training-es:start-0-0
  • German: /training-de:start-0-0
  • French: /training-fr:start-0-0
  • Portuguese: /training-pt-br:start-0-0
  • Vietnamese: /training-vi:start-0-0
  • Russian: /training-ru:start-0-0
  • Arabic: /training-ar:start-0-0

Learning Path 1: Quick Start (30 minutes)

  1. /campaign:plan "Your campaign"
  2. /content:good "Your content"
  3. /cro:page "Your landing page"

Learning Path 2: Full Training (5-6 hours)

  1. /training:start-0-0

Learning Path 3: Skill-Specific (15-30 minutes each)

  • Improve conversions: /cro:page, /cro:form, /marketing:psychology
  • Write better copy: /content:good, /content:editing
  • Launch a product: /growth:launch, /campaign:plan
  • Optimize pricing: /pricing:strategy
  • Scale SEO: /seo:programmatic, /seo:schema
  • Design referrals: /growth:referral
  • A/B testing: /test:ab-setup

Learning Path 4: CRO Mastery (60 minutes)

  1. /training:start-3-1 (CRO Fundamentals)
  2. /training:start-3-2 (Form and Signup Optimization)
  3. /training:start-3-3 (Popup and Onboarding CRO)

Parameters

platform

  • Meaning: Target AI assistant platform for npx installation.
  • Required: No. Options include claude, cursor, windsurf, cline, copilot, and all.

component

  • Meaning: Which component to install via plugin marketplace (full suite, skills only, agents only, commands only).
  • Required: No. Defaults to full suite.

language prefix

  • Meaning: Prefix for training commands to select language (e.g., /training:, /training-zh:, /training-ja:).
  • Required: Yes for training commands.

training module and lesson

  • Meaning: Specific training lesson identifier (e.g., start-0-0, start-1-1, start-2-1, start-3-1).
  • Required: Yes for training commands.

command arguments

  • Meaning: Text arguments for slash commands (e.g., campaign name, target audience, keyword).
  • Required: Varies per command.

Constraints and Caveats

  • Works only with AI assistants that support agents and skills (Claude Code, Cursor, GitHub Copilot, Windsurf, Cline).
  • Plugin marketplace installation is recommended for Claude Code only.
  • npx installation requires Node.js and npm.
  • Manual installation requires knowledge of Claude config directory (~/.claude/).
  • Training is 5-6 hours total; individual modules vary in duration.
  • Training content is identical across languages; only the command prefix changes.
  • MCP integrations require real data services and configuration (see data-reliability-rules.md).
  • The compounding effect claims that the first campaign takes 40 hours, the 5th campaign takes 15 hours (62% faster), and the 10th campaign takes 10 hours (75% faster).
  • Open source under MIT license; contributions welcome via CONTRIBUTING.md.
  • Star history is tracked on GitHub.

Failure Modes

  • Plugin installation fails if Claude Code is not properly configured or restarted.
  • npx installation fails if Node.js or npm is not installed or network issues occur.
  • Manual installation fails if paths are incorrect or files are missing.
  • Commands may not work if the AI assistant does not support the specific command syntax.
  • Training commands may fail if the language prefix is incorrect or the lesson identifier is invalid.
  • MCP integrations fail if data services are not configured or credentials are missing.
  • Clone and use method fails if git is not installed or the repository is inaccessible.

Prerequisites

  • An AI assistant that supports agents and skills (Claude Code, Cursor, GitHub Copilot, Windsurf, or Cline).
  • For Claude Code Plugin Marketplace: Claude Code installed and configured.
  • For npx installation: Node.js and npm installed.
  • For clone and use: git installed.
  • For manual installation: knowledge of Claude config directory (~/.claude/).
  • For training: compatible AI assistant (Claude Code, Cursor, etc.).
  • For MCP integrations: access to respective data services and configuration.

Examples

Campaign Launch

Use /research:market, /competitor:deep, and /campaign:plan for research and planning. Then use /content:landing, /content:email, and /content:blog for content generation. Then use /cro:page and /seo:optimize for optimization.

Content Creation

Use /content:good to generate a blog post, /content:editing to polish a draft, and /seo:keywords for keyword research.

Conversion Optimization

Use /cro:page for a homepage audit, /cro:form for lead capture optimization, /cro:signup for registration flow, and /test:ab-setup for headline A/B testing.

Growth and Strategy

Use /marketing:ideas for SaaS strategies, /marketing:psychology for pricing objections, /growth:launch for Product Hunt strategy, and /pricing:strategy for tier structure.

Training

Use /training:start-0-0 (English) or /training-zh:start-0-0 (Chinese) to begin interactive lessons.

Learning Paths

Quick Start (30 minutes) uses /campaign:plan, /content:good, and /cro:page. Full Training (5-6 hours) starts with /training:start-0-0. Skill-Specific uses targeted commands like /cro:page for conversions. CRO Mastery (60 minutes) uses /training:start-3-1, /training:start-3-2, and /training:start-3-3.

Additional Details

Detailed Slash Command Categories

The 93 slash commands are distributed across these categories:

  • campaign: Commands for planning, executing, and analyzing marketing campaigns. Example: /campaign:plan.
  • content: Commands for creating, editing, and optimizing content. Examples: /content:blog, /content:landing, /content:email.
  • SEO: Commands for keyword research, on-page optimization, and technical SEO. Examples: /seo:keywords, /seo:optimize, /seo:programmatic, /seo:schema.
  • CRO: Commands for conversion rate optimization. Examples: /cro:page, /cro:form, /cro:signup.
  • growth: Commands for growth strategies and launches. Examples: /growth:launch, /growth:referral.
  • email: Commands for email marketing sequences.
  • analytics: Commands for data analysis and reporting.
  • sales: Commands for sales enablement and lead qualification.
  • research: Commands for market and competitor research. Examples: /research:market, /competitor:deep.
  • marketing: Commands for marketing psychology and ideas. Examples: /marketing:ideas, /marketing:psychology.
  • testing: Commands for A/B testing. Example: /test:ab-setup.

Detailed Training Module Breakdown

Each training module contains multiple lessons:

  • Module 0: Getting Started (30 minutes): Introduction to the toolkit, setting up the practice project, and understanding the Markit agency scenario.
  • Module 1: Core Concepts (90 minutes): Covers marketing fundamentals, AI-assisted campaign planning, content strategy, and SEO basics.
  • Module 2: Advanced Applications (120 minutes): Deep dives into programmatic SEO, email sequences, launch strategies, and pricing optimization.
  • Module 3: CRO and Conversion (60 minutes): Focuses on conversion rate optimization techniques including page CRO, form CRO, signup flow CRO, and A/B testing.
  • Bonus content: Pattern library for reusable marketing templates and the 10x Marketer Framework for scaling efforts.

Detailed MCP Integration Setup

Each MCP integration requires specific configuration steps:

  • SensorTower: Requires API key for app store data.
  • Google Search Console: Requires OAuth credentials and site verification.
  • Google Analytics: Requires OAuth credentials and property ID.
  • Semrush: Requires API key for SEO data.
  • DataForSEO: Requires API key for SERP data.
  • Meta Ads: Requires access token and ad account ID.
  • HubSpot: Requires API key for CRM data.

All MCP integrations are configured in the data-reliability-rules.md file, which defines data freshness, accuracy, and fallback rules.

Detailed Compounding Effect

The compounding effect describes how campaign efficiency improves with experience:

  • First campaign: 40 hours to complete.
  • Fifth campaign: 15 hours (62% faster than the first).
  • Tenth campaign: 10 hours (75% faster than the first).

This improvement comes from reusing workflows, templates, and learned strategies across campaigns.

Detailed Open Source Contribution

The toolkit is open source under the MIT license. Contributions are managed through:

  • CONTRIBUTING.md: Guidelines for submitting issues, feature requests, and pull requests.
  • marketplace.json: Configuration for self-hosted marketplace deployments.
  • GitHub star history: Tracked on the repository page for community engagement.

Detailed Failure Mode Recovery

For each failure mode, specific recovery steps are recommended:

  • Plugin installation failure: Verify Claude Code is running the latest version. Restart the IDE after installation. Check network connectivity to the marketplace.
  • npx installation failure: Ensure Node.js version 18 or higher is installed. Run npm cache clean --force if network issues persist. Verify npm registry access.
  • Manual installation failure: Confirm the ~/.claude/ directory exists. Check file permissions. Verify the cloned repository is complete.
  • Command syntax issues: Ensure the AI assistant supports slash commands. Check for typos in command names. Verify arguments are enclosed in quotes.
  • Training command failure: Confirm the language prefix matches the installed training module. Use the correct lesson identifier format (e.g., start-0-0).
  • MCP integration failure: Verify API keys and credentials are valid. Check service status for outages. Review data-reliability-rules.md for fallback procedures.
  • Clone and use failure: Ensure git is installed and authenticated. Verify the repository URL is accessible. Check for disk space issues.

Related kits