Hey Dev.to! π
Just launched my automation toolkit after 6 months of development. Wanted to share with this awesome community!
The Problem
I had 30+ automation scripts scattered across folders with:
- Inconsistent CLI patterns
- No error handling
- Zero documentation
- Impossible to chain together
The Solution
Built a professional toolkit with 6 production-ready tools + FastAPI orchestrator:
π§ The 6 Tools
-
Cookie Analysis Suite (466 lines)
- Forensic browser cookie analysis
- 4-stage pipeline: parse β enrich β scan β report
- Privacy tracking detection
-
Video Enhancement Suite (1,112 lines)
- Multi-backend support (Topaz/FFmpeg/HandBrake)
- Priority-based job queue
- Concurrent processing
-
PathPulse (1,205 lines)
- Real-time file system threat detection
- 6 threat pattern detectors
- Ransomware detection
-
Windows Feature Manager (1,418 lines)
- Enable/disable Windows features
- PowerShell/DISM integration
- Backup/restore with rollback
-
Web Automation Framework (1,121 lines)
- Multi-browser Selenium automation
- 8 action types
- Headless mode + proxy support
-
ADB Automation Framework (1,555 lines)
- Android device management
- App control + input simulation
- Screen capture + file transfer
ποΈ The Orchestrator
Built with FastAPI to tie everything together:
- REST API - Execute any tool programmatically
- Web Dashboard - Visual monitoring
- Pipeline Builder - Chain tools together
- Audit Trail - Complete logging
- CI/CD Ready - GitHub Actions included
Tech Stack
- Python 3.11
- FastAPI + Pydantic
- Selenium, watchdog, subprocess
- 6,877 lines of production code
- Cross-platform (Win/Mac/Linux)
Architecture Pattern
Every tool follows the same pattern:
Models β Business Logic β CLI β API β Dashboard
This makes them easy to maintain, extend, and integrate.
Quick Start
git clone https://github.com/dopamin3fiends/ml-systems-portfolio.git
cd ml-systems-portfolio
pip install -r requirements.txt
# Test a tool
cd tools/cookie_analysis
python cli.py demo
# Start orchestrator
cd ../../
python src/backend/orchestrator_api.py
Open browser: http://localhost:8000
Lessons Learned
- Architecture matters - Even for "just scripts"
- Documentation is not optional - Future you will thank present you
- Demo modes are essential - Safe testing without real data
- Integration is hard - REST API makes it manageable
- Professional != complicated - Clear patterns beat clever code
Open Source + Support Model
The code is fully open source (MIT license) on GitHub. I'm also offering a packaged version on Gumroad with:
- Pre-configured setup
- Professional support
- Lifetime updates
- Commercial license clarity
Think: Linux is free, Red Hat makes billions. You're paying for packaging and peace of mind.
π Launch special: Use code LAUNCH20 for 20% off
Links
- GitHub: https://github.com/dopamin3fiends/ml-systems-portfolio
- Gumroad: https://dopaminefiends.gumroad.com/l/devtools
Questions?
Happy to answer anything about:
- The architecture decisions
- FastAPI integration patterns
- Building production-grade automation tools
- Open source + paid model
Thanks for reading! π
Top comments (0)