AIToday
Large Language ModelsAI Coding AssistantsZenn AI/MLPublished: Oct 8, 2026, 10:00 JST

Woodlot: free Mac app runs parallel Claude Code sessions

Woodlot: free Mac app runs parallel Claude Code sessions

3 Key Points

  1. What happened

    The developer built Woodlot, a free app that gives each new Claude Code session its own worktree and branch, pairs sessions as cards in two columns, and shows status by color and motion.

  2. Why it matters

    Running many terminal sessions at once made it hard to tell which was which or waiting for a reply, and working in one repo risked two sessions rewriting the same files at once.

WHO IT HITSDevelopers who run several Claude Code sessions locally — especially those using Bedrock or API keys, who cannot use web-based cloud execution — get a way to keep parallel work separated and see which session needs their reply.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

Context & Analysis

The starting complaint is a practical one: Claude Code in a terminal is hard to read and hard to type into, and one session is tolerable while several are not. Because Claude Code works on its own for minutes to tens of minutes after an instruction, the natural move is to start another session, and that is where the trouble begins.

The developer describes two problems. The first is that sessions blur together — you cannot recall what a session was for or whether it is waiting on you, so you end up asking the AI what is going on. Terminal tabs add their own failures: a session stopped on a multiple-choice question sits unnoticed, every tab is named claude, and two sessions in the same repository start rewriting the same files. The second is understanding debt: handing implementation through to a pull request takes your hands off, but leaves you unable to follow what was built, which becomes acute when the AI asks you to make a judgment call.

Woodlot's answer is to combine git worktree management, a view of all sessions, and reading material for catching up in one app. The catch-up feature condenses what happened in your absence into a single page covering the problem, the options weighed, the choice made, and the implementation, so you can return to the decision. Other features were added one at a time through daily use: handover that spins up a fresh worktree from a new one-line instruction, read-only sub-sessions branched off the current conversation, a four-stage track for changes (local, committed, unpushed, PR), CI watching via GitHub, a loop that picks up assigned issues and runs them to merge, a display of context weight and the 5-hour and weekly usage limits, and per-repository settings for the AI account, GitHub account, browser profile, and a setup command run right after a worktree is created.

FAQ
What does Woodlot cost?
It is free. You download the DMG from woodlot.app and move Woodlot.app into /Applications.
What do I need to run it?
An Apple Silicon Mac with macOS 12 or later, the claude CLI with a working setup (such as a Claude login or Bedrock), and git. gh, Chrome and the Claude in Chrome extension are optional.
Does Woodlot send my code or conversations anywhere?
It only launches the Claude Code already on your machine, and conversation records stay under ~/.woodlot on your Mac. It does not read your Claude login credentials.
How does it help when I have fallen behind on a delegated task?
Its catch-up feature has the AI summarize what happened in about five minutes of reading — the problem, the options considered, what was chosen, and how it was implemented.

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleAI cluster networking reading list targets GPU engineers