From 25 Aliases to One Command: Building a Smart CLI Workspace Launcher
How I replaced dozens of hardcoded shell aliases with a fuzzy-searching workspace launcher that actually works
The QL Journey Series
Follow the evolution of Quick Launch (ql) from a simple project launcher to a full AI-powered development environment. Real code, real decisions, real lessons learned.
Part 1
Part 6
Implementing JetBrains' Observation Masking: 80% Context Reduction for AI Agents
AI coding agents hit context limits fast. JetBrains' research shows simple observation masking beats complex summarization—here's how to implement it.
Part 3
Coming Soon!
Check back soon for the next part of this series.
Part 4
Coming Soon!
Check back soon for the next part of this series.
Part 5
Coming Soon!
Check back soon for the next part of this series.
Part 6
Coming Soon!
Check back soon for the next part of this series.
Part 7
Coming Soon!
Check back soon for the next part of this series.
Part 8
Coming Soon!
Check back soon for the next part of this series.
Part 9
Coming Soon!
Check back soon for the next part of this series.
Part 10
Coming Soon!
Check back soon for the next part of this series.
Part 11
Coming Soon!
Check back soon for the next part of this series.
About This Series
This series documents the complete journey of building Quick Launch (QL)—a Python-based development toolkit that started as “I’m tired of typing these aliases” and evolved into an AI chat system with semantic search, automatic backups, and intelligent context management.
What makes this unique: You’re seeing the ACTUAL evolution, warts and all. Some parts are published out of order because that’s how development works—you build what’s needed, when it’s needed.
What You’ll Learn
- Part 1: From 25 Aliases to One Command - The project launcher that started it all
- Part 2: Never Lose an AI Conversation Again (coming soon) - Building the backup system
- Part 3: The Always-On Watcher (coming soon) - Real-time conversation tracking
- Part 4: Searching 10,000 Messages in 0.2s (coming soon) - Qdrant semantic search
- Part 5: The Chat Mentor (coming soon) - Q&A over conversation history
- Part 6: Implementing JetBrains’ Observation Masking - 80% context reduction for long chats
- Parts 7-10: Multi-agent crew system, time tracking, blog integration, and more (coming soon)
Status
🚧 In Progress - Parts 1 & 6 published. Building the rest as the system grows.
Why Non-Sequential?
Unlike typical tutorial series, this reflects how real software evolves:
- Part 1 was built first (Oct 2024)
- Part 6 was needed urgently (Dec 2024) when context overflow became critical
- Parts 2-5 will fill in the story as I write them
This is how actual development works—solving the most pressing problems first, documenting them second.
The project is open source: github.com/NightOwlCoder/quick-launch