Worktrunk is live · free · last verified 8 Nov 2026 · viability monitor
Worktrunk
Worktrunk is a free, MIT/Apache-licensed CLI that makes Git worktrees as easy as branches for developers running parallel AI agents.
For developers already running parallel AI agents, Worktrunk pays for itself in the first afternoon: hooks, LLM commit messages and copy-on-write cache sharing on APFS/btrfs/XFS are daily wins, and wt switch -x claude -c feature-a -- 'Add user authentication' collapses a three-step git worktree add / cd / claude dance into one line. It is not a replacement for Fork or GitKraken and won't teach you worktrees — it's a scripting-grade layer over git worktree. Adopt it when worktree count, not branch count, is your bottleneck.
Verdict generatedVerified 6d ago · liveness 62/100 · cite: rightaichoice.com/tools/worktrunk
- Developers running 5–10+ AI agents (Claude Code, Codex) in parallel across feature branches
- Trunk-based teams with frequent worktree creation, merge and cleanup churn
- Git power users who want worktrees addressed by branch name instead of filesystem path
- Engineers automating local setup with hooks (install deps, start dev servers per worktree)
- Beginners who haven't used Git branches or worktrees before — this is not a teaching tool
- Developers working a single branch at a time, where parallel isolation adds no value
- Anyone who prefers a graphical Git client over a terminal CLI
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Worktrunk if you work one branch at a time or haven't yet used Git worktrees — the branch-name addressing, hooks and cache sharing only pay off once worktree churn is your bottleneck.
Copy-on-write cache sharing needs APFS, btrfs or XFS — on other filesystems each worktree rebuilds target/ and node_modules/, which can mean a long cold start per branch.
Worktrunk is free and open source under MIT OR Apache-2.0 — there is no paid tier, no seat count and no usage cap to compare. The real comparison is effort, not dollars: against plain git worktree add/remove commands you save typing and per-worktree setup, and against a general Git GUI you gain branch-name addressing, hooks and CoW cache sharing but give up graphical history, blame and conflict resolution.
In short
Worktrunk — Worktrunk is a free, MIT/Apache-licensed CLI that makes Git worktrees as easy as branches for developers running parallel AI agents. Best for Developers running 5–10+ AI agents (Claude Code, Codex) in parallel across feature branches, Trunk-based teams with frequent worktree creation, merge and cleanup churn, Git power users who want worktrees addressed by branch name instead of filesystem path. Free to use.
What is Worktrunk?
Worktrunk — Worktrunk is a free, MIT/Apache-licensed CLI that makes Git worktrees as easy as branches for developers running parallel AI agents. Best for Developers running 5–10+ AI agents (Claude Code, Codex) in parallel across feature branches, Trunk-based teams with frequent worktree creation, merge and cleanup churn, Git power users who want worktrees addressed by branch name instead of filesystem path. Free to use.
Worktrunk is a CLI for Git worktree management, built for developers running parallel AI agent workflows. Git's native worktree feature gives each agent its own working directory so they don't step on each other's changes, but the UX is clunky: even starting a new worktree means typing git worktree add -b feat ../repo.feat and then cd ../repo.feat. Worktrunk replaces that with wt switch -c feat. Worktrees are addressed by branch name and paths are computed from a configurable template, so the mental model matches ordinary Git branching. The core commands are deliberately small: wt switch to move between worktrees, wt list for branch status (with @ marking the current worktree, + for staged changes, and ↑/⇡ for ahead and unpushed commits), wt merge to squash, rebase onto main, fast-forward and clean up in one step, and wt remove to delete a worktree and its branch together. Around those sit hooks that run on create, pre-merge and post-merge; LLM-generated commit messages from diffs; an interactive picker with streaming CI status and diff, log, PR and comment previews; and a hash_port template filter that gives each worktree a unique dev server port. For anyone juggling 5–10+ agents, copy-on-write build cache sharing on APFS, btrfs and XFS lets ten worktrees share target/ and node_modules/ without rebuilding or copying them. PR checkout is wt switch pr:123, and multi-agent launch is wt switch -x claude -c feature-a -- 'Add user authentication'. Install paths cover Homebrew, Cargo, WinGet, Arch Linux and Conda/Pixi. It is not a Git GUI — it assumes you already know branches and worktrees and are working trunk-based with frequent parallel branches.
Key features
- Switch between worktrees by branch name with wt switch feat
- Create and switch into a new worktree in one command: wt switch -c feat
- Launch an AI agent after switching: wt switch -x claude -c feature-a -- 'Add user authentication'
- wt list with @ current-worktree marker, + staged changes, ↑ ahead and ⇡ unpushed counts
- wt list --full adds CI status and AI-generated summaries per branch
- wt merge: squash, rebase onto main, fast-forward merge and cleanup in one command
- wt remove deletes the worktree and its branch together
- Hooks that run commands on create, pre-merge and post-merge
- LLM-generated commit messages from diffs (wt step commit)
- Interactive picker with streaming CI status and diff, log, PR and comment previews
- Copy-on-write build cache sharing for target/ and node_modules/ on APFS, btrfs and XFS
- PR checkout: wt switch pr:123 jumps straight to a PR's branch
- hash_port template filter assigns each worktree a unique dev server port
- Aliases and per-branch variables for custom wt <name> commands and hook templates
- Shell integration so Worktrunk commands change directories automatically
Behind the verdict
Worktrunk attacks a real and recent pain: AI coding agents can now run for long stretches unsupervised, so managing 5–10+ in parallel is practical, and each one needs its own working directory to avoid clobbering another's changes. Git worktrees solve the isolation, but the command surface is hostile — branch names get typed three times, you cd by raw filesystem path, and cleanup is a two-step worktree-remove plus branch-delete. Worktrunk's pitch is that worktrees should be addressed the way branches are, by name, with paths computed from a configurable template. That single design decision is what makes wt switch feat readable and wt switch -c feat a one-line replacement for the git worktree add -b dance. The command set is intentionally thin: switch, list, merge, remove. What earns the tool its place are the surrounding details. Hooks fire on create, pre-merge and post-merge, so install-deps or start-dev-server setup happens automatically per worktree. wt list shows more than paths — the @ marks your current worktree, + flags staged changes, and ↑/⇡ indicate commits ahead of main and unpushed commits, with --full adding CI status and AI-generated per-branch summaries. wt merge does squash, rebase onto main, fast-forward and cleanup as one operation, printing each stage as it goes. wt remove deletes the worktree and its branch together rather than making you chain two commands. The interactive picker streams CI status alongside diff, log, PR and comment previews. The feature that matters most at scale is easy to miss: copy-on-write build cache sharing. On APFS, btrfs and XFS, ten worktrees share target/ and node_modules/ without rebuilding or copying, which removes the cold-start tax that normally makes parallel worktrees expensive in Rust, Node or anything with a heavy dependency tree. Alongside that, hash_port assigns each worktree a unique dev server port, noPort collisions between parallel branches. LLM-generated commit messages from diffs, aliases with per-branch variables for custom wt <name> commands, and shell integration that lets commands change directories round out the ergonomics layer. Where it doesn't fit: this is a terminal CLI and assumes fluency with branches and worktrees. Nothing here teaches Git, and there is no GUI — no history browsing, blame or conflict-resolution UI. It assumes trunk-based work with frequent parallel branches; long-lived GitFlow release branches fight the model. It also manages worktrees for agents rather than orchestrating them — it doesn't assign tasks or coordinate between agents, it just gives each one a clean isolated directory and fast commands to get in and out. If you work one branch at a time, none of this earns its keep. If you're a Git power user with 5–10 agents in flight and a heavy build cache, the combination of branch-name addressing, CoW sharing and one-command merge/remove is the reason to adopt.
Models under the hood
as of 2026-11-14
Limitations
- Worktrunk is a terminal CLI and requires proficiency with Git branches and worktrees — it won't teach you either.
- It provides isolated directories for parallel agents rather than orchestrating task assignment or coordination between them; you still decide what each agent does and when.
- It is not a Git GUI: there is no history browsing, blame, or conflict-resolution UI, so pair it with your editor's Git tooling or a client like Fork or GitKraken.
- Copy-on-write build cache sharing only works on filesystems that support it — APFS, btrfs and XFS — so other setups rebuild per worktree.
- The workflow assumes trunk-based development with frequent parallel branches; long-lived release-branch models don't fit.
as of 2026-10-08
Researching Worktrunk? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Is Worktrunk still active?
Worktrunk is live · free · last verified 8 Nov 2026 · viability monitor.
Viability score
How well maintained and how widely used is Worktrunk? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: November 2026
How we score →What do users say about Worktrunk?
What people actually say about Worktrunk — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
ResearchedDiscussed by users across 4 sources (Hacker News, YouTube, Bluesky, GitHub) · researched Nov 14, 2026.
Weighted by the 71 posts each of 4 sources contributed.
- +One-command worktree creation and switching: `wt switch -c feat` replaces a three-step git incantation
- +Branch-name addressing for worktrees matches how developers already think about Git
- +Copy-on-write caching of `target/` and `node_modules/` eliminates per-worktree rebuilds and disk bloat
- +Hooks on create, pre-merge and post-merge let teams standardize per-worktree setup
- +`wt merge` collapses squash, rebase, fast-forward and cleanup into one command
- −Agents sometimes create worktrees inside the repo, breaking the adjacent layout users rely on
- −Advanced, CLI-only design assumes deep Git worktree knowledge — no GUI for newcomers
- −Disk-space and worktree-mess concerns persist for users without CoW-capable filesystems
- −Port and env isolation depends on manual hook/template configuration, not defaults
- −Not a fit for GitFlow or single-branch workflows that never needed worktrees
- • LLM-generated commit messages and summaries consume your own API credits — not Worktrunk's cost, but a real per-use expense
- • Ongoing maintenance time to author hooks and env/port templates for each stack
- • Disk savings from CoW only materialize on APFS/btrfs/XFS — ext4 users pay in storage
How we verified this
Every listed fact about Worktrunk — pricing, features, links, status — is re-read from the vendor's own pages by our verification pipeline and each change is logged with a date; the last pass ran on 8 Nov 2026. Analysis is drafted with AI models and checked against our editorial standards before publishing; see how we verify.
Verification history
We have re-verified Worktrunk 14 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — changed SSL certificate
- — changed setup time, who should skip it, pricing tiers, switching out, features, who it suits, hidden costs, summary, limitations, switching in, use cases, our verdict, pricing fit, description
- — changed limitations, pricing page
- — changed description, who it suits, features, who should skip it, pricing tiers, summary, our verdict
- — changed features, who it suits, description, our verdict, hidden costs, limitations, switching in, switching out, pricing fit, setup time, who should skip it, summary, use cases, pricing tiers, pricing page
- — changed domain expiry
Showing the 6 most recent of 14 verification passes.
Free to cite with attribution — this page re-verifies continuously.
How much does Worktrunk cost in 2026?
Worktrunk: Free to use. Pricing last verified 8 Nov 2026.
Plans compared
Every published Worktrunk tier, its price as the vendor publishes it, who it fits, and what it adds over the tier below. The 12-month cost projection follows.
| Plan | Price | Billing | Who it's for |
|---|---|---|---|
| Free | Free | — | Individual developers and teams running parallel AI agent workflows who want worktree management without a license fee — the whole CLI ships under MIT OR Apache-2.0.Adds: Free entry point covering the full CLI: wt switch, wt list, wt merge, wt remove, hooks, LLM commit messages, interactive picker, CoW cache sharing and PR checkout. |
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Where the pricing makes sense
The company stage and team size where Worktrunk's pricing actually pencils out — and where peers do it cheaper.
Worktrunk is free and open source under MIT OR Apache-2.0 — there is no paid tier, no seat count and no usage cap to compare. The real comparison is effort, not dollars: against plain git worktree add/remove commands you save typing and per-worktree setup, and against a general Git GUI you gain branch-name addressing, hooks and CoW cache sharing but give up graphical history, blame and conflict resolution.
Has Worktrunk's pricing changed?
No plan price for Worktrunk has moved since we began logging in Aug 2026; the pricing page is re-read on every verification pass.
Who should pick Worktrunk?
Pick Worktrunk if you are Developers running 5–10+ AI agents (Claude Code, Codex) in parallel across feature branches, Trunk-based teams with frequent worktree creation, merge and cleanup churn, Git power users who want worktrees addressed by branch name instead of filesystem path, Engineers automating local setup with hooks (install deps, start dev servers per worktree).
Look elsewhere if you are Beginners who haven't used Git branches or worktrees before — this is not a teaching tool, Developers working a single branch at a time, where parallel isolation adds no value, Anyone who prefers a graphical Git client over a terminal CLI, Teams on GitFlow-style long-lived release branches that clash with trunk-based assumptions.
Skip it if Skip Worktrunk if you work one branch at a time or haven't yet used Git worktrees — the branch-name addressing, hooks and cache sharing only pay off once worktree churn is your bottleneck.
GeneratedReal-world workflow fit
Concrete scenarios for the personas Worktrunk actually fits — and what changes day-one when you adopt it.
You spin up one worktree per task with wt switch -x claude -c feature-a -- 'Add user authentication', then two more for the pagination bug and API tests, each launching its agent on creation with a first prompt.
Outcome: Three agents work in isolated directories, wt list shows each branch's ahead/unpushed counts, and cache sharing on APFS means target/ and node_modules/ aren't rebuilt for every branch.
Instead of stashing your work, you run wt switch pr:123 to jump directly to the PR's branch in its own worktree, with the interactive picker streaming CI status, diff, log and comment previews.
Outcome: You review in a clean directory without touching your working branch, then wt remove when the PR is merged.
After finishing work in a worktree, you run wt merge main, which generates and commits a message from the diff, rebases onto main, fast-forwards, and removes the worktree and branch in the background.
Outcome: One command replaces commit, rebase, merge and cleanup, and wt list confirms main is up to date with the worktree gone.
Use cases
- Run multiple Claude Code agents in parallel, each in its own isolated worktree, and clean up in one command.
- Manage feature branches with a simple wt switch instead of git worktree add plus cd.
- Automate per-worktree setup (install deps, start dev server) with post-start hooks.
- Generate LLM commit messages from diffs to speed up the PR process.
- Squash-merge a branch and remove the worktree in one wt merge command.
- Jump straight to a PR's branch with wt switch pr:123 for faster code reviews.
- Share target/ and node_modules/ between worktrees to avoid rebuilding per branch.
- Give each worktree a unique dev server port with the hash_port filter so parallel branches don't collide.
Setup time & first value
How long it actually takes to get something useful out of Worktrunk — broken out by persona, not the marketing-page minute.
Most developers are productive within minutes: install via brew install worktrunk (or Cargo, WinGet, Arch, Conda/Pixi), run wt config shell install for directory-changing shell integration, then wt switch -c your-first-feature. Hooks and LLM commit messages take longer because they need per-repo configuration and an LLM set up, but the core switch/list/merge/remove loop works immediately.
Switching to or from Worktrunk
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From plain git worktree add/remove: keep your branch habits and replace the add -b plus cd dance with wt switch -c feat, and the two-step remove with wt remove.
- →From manual cd ../repo.branch navigation: install shell integration with wt config shell install so wt switch changes directories for you.
- →From per-branch setup scripts: move install-deps and start-dev-server steps into Worktrunk create and post-start hooks.
- →From a PR checkout alias: use wt switch pr:123 to check out a PR branch into its own worktree in one command.
- ↗To plain git worktree: your worktrees are ordinary Git worktrees, so git worktree list, git worktree remove and git branch -d keep working on the same directories.
- ↗To a general Git GUI (Fork, GitKraken): Worktrunk has no history, blame or conflict-resolution UI, so you'd move those tasks to the GUI while keeping worktree creation in the terminal.
Resources & guides
Tutorials & Learning
YouTube returned 6 videos for “Worktrunk”, and we withheld 6: 6 could not be judged, because “Worktrunk” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Worktrunk.
Official links
What are the best Worktrunk alternatives?
The closest alternatives to Worktrunk we list are Zed, Trae and ReadMe. Ranked by direct product-type match, not category overlap.
| Alternative | Pricing | Starts at | Last verified |
|---|---|---|---|
| Zed | Freemium | Pricing not published | — |
| Trae | Freemium | Pricing not published | — |
| ReadMe | Freemium | Pricing not published | — |
Alternatives to Worktrunk
View all
Tools that pair well with Worktrunk
Common stack mates teams adopt alongside Worktrunk, with the specific reason each pairing earns its keep.
Zed
Zed is a GPU-rendered Rust code editor that runs parallel AI agents at native speed
Trae
Trae is an AI coding IDE with parallel SOLO cloud agents and usage-metered pricing
ReadMe
ReadMe turns an OpenAPI spec into interactive API docs with AI linting, Git-synced review branches, and an MCP server agents can query.
Worktrunk vs its closest rival
Worktrunk's closest head-to-head is with Presto Voice.
Our Worktrunk vs Presto Voice verdict written Jul 3, 2026 is more than 90 days old, so it is being re-reviewed against current data and is not shown.
Frequently asked questions
Categories
Best-of guides
For makers of Worktrunk
- Get a Verified Listing for Worktrunk ($39): a written decision within 1 working day.
- Free listing for another tool: Free reviews are decided in order, within 3 weeks of the review starting.
Used Worktrunk? Help shape our editorial sentiment research.