I kept running into a structural problem with AI coding tools.
You try one agent. It makes changes. You try another. Now you're manually diffing branches and cleaning up experimental commits.
The tools are powerful, but the workflow isn’t designed for comparison.
So I built Parallel Code.
It runs Claude Code, Codex, and Gemini CLI natively (no API wrapper, no feature abstraction), but isolates each agent in:
its own terminal
its own Git worktree
its own feature branch
The workflow becomes:
Spawn N worktrees
Assign one AI agent per worktree
Let them implement independently
Compare diffs
Merge the best result
Instead of sequential trial-and-error, you get parallel exploration with safe isolation.
Why not just use tmux?
You can — but you still manually manage:
branch creation
worktree cleanup
session naming
context switching
Parallel Code automates the Git structure and keeps the sessions visually organized.
This is aimed at people already using AI CLIs heavily.
I’m curious: Has anyone else moved toward multi-agent parallel workflows instead of single-agent iteration?
GitHub: https://github.com/johannesjo/parallel-code
Show HN: Parallel Code – Running multiple AI agents in parallel with worktrees | Heykuki News