Built by an Indie Dev,
for Indie Devs
The story behind AgentJuggler and why it exists.
@Pawon_Pawon
Indie builder, weekend prototyper
I'm an indie developer who juggles multiple side projects on weekends. When AI coding agents like Claude Code and Codex came along, they were incredible productivity multipliers. But using them was chaos.
Every project needed its own terminal sessions. Context would evaporate when I switched between things. I'd lose track of which agent was doing what, which project was where, and what state each terminal was in. Browser tabs piled up. Workflow fell apart.
So I built AgentJuggler — first as a personal hack to manage 4-6 concurrent tasks without losing my sanity. One workspace where every project has its own set of AI-powered and plain terminals, each opening right in the project directory, each remembering where I left off.
It worked so well for me that I decided to share it. I've been building it in public on X, posting raw screenshots, wins, and progress. No polish, no marketing spin — just a tool that does what it says.
How We Think About Software
Local-First
Your data stays on your machine. No cloud accounts, no syncing, no servers to depend on. JSON files you can read and back up yourself.
No Telemetry
AgentJuggler doesn't phone home. No analytics, no crash reports, no tracking. It's a local tool — it stays local.
Bring Your Own Keys
We don't proxy API calls. Your Claude and Codex API keys go directly to their respective services. We never see or store them.
Simple Architecture
Python FastAPI backend, vanilla JS frontend, real PTY terminals. No Electron bloat, no framework jungle. One binary to install.
Building in Public
Follow the journey on X. Raw progress updates, screenshots, feature drops, and honest takes on what's working and what's not.
Follow @Pawon_Pawon