MarketingFlow
Marketing Strategy Execution Platform
Transform expert marketing knowledge into actionable strategies
Overview
MarketingFlow is a cross-platform Flutter application that runs on iOS, Android, macOS, Windows, Linux, and Web. It extracts the complete knowledge base from the ai-marketing-skills open-source project and transforms it into an interactive strategy execution tool.
75 marketing strategies across 11 categories — from content operations to sales playbooks — are packaged into a single codebase that works on every platform where users can browse, search, and execute strategies powered by LLM.
Features
Complete Knowledge Extraction
All marketing assets from the source repository are extracted and categorized:
| Type | Count | Description |
|---|---|---|
| Strategy Definitions | 11 | Core SKILL.md workflow definitions |
| Expert Personas | 9 | Specialized evaluator profiles (LinkedIn, SEO, Newsletter...) |
| Automation Scripts | 42 | Python scripts for marketing automation |
| References | 13 | Templates, patterns, and guidelines |
| Scoring Rubrics | 5 | Content quality evaluation frameworks |
| Documentation | 13 | Category README files |
| Configurations | 10 | Environment templates and configs |
| Dependencies | 11 | Requirements files |
| Total | 114 |
13 Marketing Categories
| Category | Assets | Key Tools |
|---|---|---|
| Content Ops | 26 | Expert panel evaluation, platform-specific content (LinkedIn, Instagram, X, YouTube Shorts, Newsletter), humanizer, SEO strategy |
| Finance Ops | 13 | CFO analyzer, scenario modeler, ROI calculator, QuickBooks integration |
| Outbound Engine | 13 | Cold outbound sequences, competitive monitoring, lead pipeline, Instantly audit |
| Sales Pipeline | 11 | Deal resurrector, ICP learning, RB2B webhook/routing, trigger prospecting |
| SEO Ops | 8 | Content attack briefs, Google Search Console integration, trend scouting |
| Growth Engine | 7 | Experiment engine (ICE scoring), weekly scorecard, pacing alerts |
| Sales Playbook | 7 | Call analyzer, value pricing briefing/packager, pricing patterns |
| Revenue Intelligence | 6 | Client reports, Gong insights, multi-touch attribution |
| Conversion Ops | 5 | CRO audit framework, survey lead magnets |
| Team Ops | 5 | Meeting action extraction, "Elon Algorithm" performance audits |
| Podcast Ops | 5 | Full pipeline from ideation to distribution |
| Telemetry | 5 | Logging, reporting, version tracking |
| Security | 3 | Pre-commit hooks, sanitizer |
App Features
- Dynamic Form Builder — Automatically generates input forms based on each strategy's required variables
- Strategy Execution — Combines system prompts with user inputs and sends to Anthropic Claude API
- Markdown Viewer — Rich rendering of generated strategies with copy-to-clipboard
- Bilingual UI — Full Korean and English interface, switchable in settings
- Warm Design System — Professional coral/amber palette designed for marketing teams, not a tech-heavy look
- License Compliance — Full MIT license and attribution displayed in-app
- Auto-Update Script — Python extractor to sync with upstream repository changes
Architecture
lib/
├── main.dart # App entry with state management
├── app_state.dart # Global state (locale, API key)
├── theme.dart # Warm coral design system
├── l10n/
│ └── app_locale.dart # EN/KR translations
├── models/
│ └── marketing_skill.dart # Data models + JSON loader
├── screens/
│ ├── home_screen.dart # Search, filter, browse
│ ├── skill_detail_screen.dart # Strategy execution
│ ├── settings_screen.dart # Language + API config
│ └── about_screen.dart # License & attribution
├── services/
│ └── ai_response_service.dart # Anthropic Claude API
└── widgets/
├── dynamic_form_builder.dart # Auto-generated input forms
└── markdown_viewer.dart # Rich markdown display
assets/
└── marketing_knowledge_base.json # 114 extracted marketing assets (1.15 MB)
scripts/
└── extract_knowledge.py # Python extractor for upstream sync
Getting Started
Prerequisites
- Flutter SDK 3.11+
- Anthropic API key (for strategy execution)
- Python 3.8+ (for knowledge base updates)
Installation
git clone https://github.com/kimdzhekhon/MarketingFlow.git
cd MarketingFlow
flutter pub get
flutter run
One-Command Setup (from scratch)
bash setup_full_extraction.sh
This script clones the source repository, extracts all 114 assets, generates the app logo, and sets up the complete Flutter project.
Update Knowledge Base
When the upstream repository adds new content:
git clone https://github.com/ericosiu/ai-marketing-skills.git /tmp/source
python3 scripts/extract_knowledge.py /tmp/source assets/marketing_knowledge_base.json
Data Pipeline
ericosiu/ai-marketing-skills (GitHub)
│
▼
scripts/extract_knowledge.py ── Recursive scan of .md, .py, .json, .env, .txt
│
▼
assets/marketing_knowledge_base.json ── 114 items, 9 types, 13 categories
│
▼
Flutter App ── Browse → Select → Input → Execute → View Results
License
This project is licensed under the MIT License.
Attribution
The marketing knowledge base is derived from ai-marketing-skills by Eric Siu / Single Grain.
MIT License — Copyright (c) 2026 Single Grain
Full license text is displayed in the app's About screen.
Built with Flutter + Claude API