0.82.02026-11-14
Release Notes
Improved
- Backups before forced removal: wt remove --force snapshots the worktree, including untracked files, to refs/wt-backup/<branch> and prints a recovery command. Ignored files, changes inside submodules, and separate staged versions are excluded. A failed backup stops removal; an unreadable status still permits removal without one. Docs (#4545, thanks @yajo for #4544)
- Codex activity hooks ship separately: skills stay in worktrunk; activity hooks move to worktrunk-markers, so skill updates don't trigger hook reapproval. (Breaking: existing installs need codex plugin add worktrunk-markers@worktrunk to retain activity markers after updating.) Docs (#4707, thanks @win5do for #4654)
- Claude Code /wt-diff pane: shows the branch diff for worktrees the session has worked in, including work detected from subagents. Failed worktree lookups are retried, and tool errors reach Claude's hook diagnostics. Docs (#4638, #4668, #4689, thanks @rudycelekli and @Yash2031v)
- Shell completion: typing wt switch origin/ completes that remote's branches. Hook names honor per-project user overrides and source-qualified filters; completion honors -C, --config, config overrides, and alias precedence over executables on PATH. PowerShell filters suggestions by the typed prefix. (#4613, #4549, #4551, #4548, #4560, #4721, thanks @Bortlesboat, and @thernstig for #4612)
- JSON survives later failures: commands retain confirmed results when a later step fails, while diagnostics go to stderr and the exit status signals failure. wt switch --format=json emits one result after execution. List schema 1 reports the tracked remote branch when its name differs locally. (#4345, #4574, #4584, #4756)
- Removal feedback: failures describe the surviving checkout and branch. --delete-branch is visible in help as the override for delete-branch = false, and the picker briefly confirms successful Alt-x removal. (#4457, #4615, #4669, thanks @Yash2031v for the picker feedback and @cazador481 for #4614)
- Fewer repeated Git operations: committing skips the second staging pass when no pre-commit hook ran; concurrent list tasks share identical ahead/behind counts. (#4540, #4663)
Fixed
- Copying ignored files: .worktreeinclude can select paths inside wholly ignored directories, such as .cache/types/. Copy and promotion also enforce destination containment before creating directories or moving files. (#4751)
- Foreground cancellation: signal handling closes races around child startup and completion, so Ctrl-C stops subsequent command steps. On Unix, wt step tether stops its supervised processes on interruption and now propagates a command's failure status. (#4596, #4609, #4623, #4598, thanks @drldavis for #4622)
- Hook execution and feedback: unselected templates no longer cause errors or consume arguments. Filtered background hooks retain independent user and project pipelines, and template failures identify the command. Background hooks running in different directories receive separate announcements. (#4429, #4781)
- Config loading: rejected environment overrides warn and fall back to file settings consistently across commands; unreadable config files are reported instead of silently ignored. System-config deprecations point to the file that needs editing. (#4358, #4564, #4767)
- State values: clearing a key removes all repeated values. Branch-local reads and clears exclude variables belonging to branches whose names extend with .vars.*. State tables preserve literal pipes, links, and formatting characters. (#4566, #4779, #4780)
- Log and terminal handling: authentication-probe output is redacted from diagnostics; Unix logs use private permissions, and permission repair leaves symlink targets unchanged. Commit subjects, PR context, and list summaries escape terminal control characters. (#4604, #4611, #4605, #4730)
- SSH remotes with a non-default port: port-less legacy project IDs no longer supply approvals or project settings to a different SSH server. Those repositories need port-qualified approvals and project overrides. (#4606)
- Removal with submodules: text-mode removal warns about submodule commits reachable from no remote; the warning does not prevent removal. Orphan cleanup preserves fsmonitor daemons serving live submodules. (#4460, #4729)
- Fork branch conflicts and GitHub Enterprise: GitLab and Azure DevOps use a prefixed, fetch-only branch when a fork's branch name conflicts locally. GitHub Enterprise PR lookup uses the remote’s host, or a supplied PR URL’s host, when gh is authenticated there. (#4542, #4710, #4722, thanks @anqorithm, and @pgilad-rna for #4709)
- Azure DevOps remotes: collection-prefixed URLs, legacy SSH URLs, and URLs omitting the project segment resolve the correct project for links and CI lookup. (#4764, thanks @owpac for #4763)
- Shell integration: directory changes preserve trailing spaces, and PowerShell treats paths containing brackets literally. Installation continues configuring other shells after an rc-file error; uninstall matches recog
0.81.02026-10-20
Release Notes
Improved
- Hooks can prompt: hooks that run serially in the foreground (pre-* hooks, wt hook <type> --foreground) inherit wt's stdin, so gum confirm 'trust this worktree?' && mise trust works as a hook. Concurrent and background hooks read EOF. (Breaking: hooks no longer receive their context as JSON on stdin; pass values as template arguments, or whole objects with the new tojson filter. wt step for-each keeps its JSON.) Docs (#3129, #4451, #4524, thanks @endigma for #3093)
- Experimental wt switch --path: places one worktree outside the worktree-path template and keeps its branch name, as in wt switch --create feature/JIRA-1234 --path ../dark-mode. Commands reach it by branch or by path. Docs (#4317, thanks @christowiz for #1982)
- wt merge --branch: merges another worktree's branch without leaving the current worktree. Commit, squash, rebase, and hooks run in the source worktree, and cleanup removes it. Docs (#4360, thanks @liuyinchu)
- OpenCode 2 worktrees go through Worktrunk: after re-running wt config plugins opencode install, worktrees OpenCode creates and removes use wt switch --create and wt remove, as the Claude Code plugin's do. OpenCode's worktree.directory stops applying, and new branches start from the default branch. Like the Claude Code plugin, it opens an existing branch's worktree, which OpenCode's remove then removes. Docs (#4489, thanks @rzane for #4488)
- Faster bulk removal: removal hands files to one background deleter per repository, not one rm -rf each, and deletes read-only directories it used to leave in .git/wt/trash. wt step prune no longer stalls when two worktrees share a directory name, as Codex worktrees do. (Breaking: prune runs pre-remove hooks concurrently; hooks writing shared state must coordinate.) (#4516, #4361, #4337)
- wt switch - returns to the worktree it left, including a detached worktree or the second checkout of a branch, and wt remove - targets the same worktree. Pre-switch hooks see the resolved destination, so wt switch origin/feature passes branch=feature. (Breaking: from a bare repository's directory, commands given no argument act on the worktree of the branch HEAD names.) (#4497, #4496, #4521)
- Template errors name the variable: {{ brnch }} reports undefined variable brnch and suggests branch. (Breaking: filters such as sanitize reject an undefined input, so {{ upstream | sanitize }} needs a guard or default on a branch with no upstream.) (#4350, #4440)
- Templates use MiniJinja 3 (Breaking): objects and lists render Python-style ({'port': 3000}) instead of as JSON, so use | tojson where JSON is needed; round rounds half to even; title capitalizes fewer word boundaries (feat/x-y becomes Feat/x-Y); division by zero errors. (#4442, #4524)
- CI status shows ⚠ for any failed forge call, where some failures, such as a Gitea 404, used to leave the cell blank. A forge CLI with no login for the host still leaves it blank, and wt config show --full names the host it checked. (#4436, #4449)
- Removal explains itself: --force lists the changes it discards, and a refusal over untracked files suggests git stash -u. A multi-target wt remove continues past a target whose pre-remove hook or removal fails, and exits 1. A branch-only removal names the detached worktree left at the branch's path. (#4347, #4357, #3791, #4313, thanks @chachi for #3769)
- The picker can open without its preview: [switch.picker] preview = false hides it until Alt-p. (#4428, thanks @emacampolo for #4072)
- Squash backups print a recovery command: squashing staged changes prints the refs/wt-backup/<branch> ref and a wt switch --create --base <sha> command that restores them in a new worktree, even after wt merge removed the original. (#4347)
- Shell integration: wt merge and wt remove suggest restarting the shell when integration is installed but inactive. On Unix, the PowerShell profile comes from the installed pwsh, so a custom XDGCONFIGHOME works. (Breaking: installing for Nushell requires nu, and for PowerShell on Unix, pwsh.) (#4343, #4422, #4436)
- System config has one path per platform: /Library/Application Support/worktrunk/config.toml on macOS, %PROGRAMDATA%\worktrunk\config.toml on Windows, /etc/xdg/worktrunk/config.toml elsewhere, or WORKTRUNKSYSTEMCONFIGPATH. (Breaking: $XDGCONFIG_DIRS is ignored and macOS no longer falls back to /etc/xdg; a file there stops loading without a warning.) (#4169)
- expectedworktreepathofbranch() template function: returns where wt switch --create would put a branch's worktree, even after the worktree at that path has moved to another branch. (#4502, thanks @BenJurewicz for #4480)
- Smaller improvements: wt switch --create reports an invalid branch name in one line instead of Git's error; hook failures name the hook's source (pre-merge user:lint failed (exit code 7)); the wt list loading footer says how many rows fit; wt step for-each labels detached worktrees by path; install-statusline keeps settings.json's key order; the Codex plugin shows the Worktrunk icon. (#4349, #4352, #4333, #4498, #4420, #4327)
Fixed
- Foreground hooks that change terminal settings: a hook that touched the terminal, such as turbo on a cache miss, was stopped by SIGTTOU and wt hung. Hooks now run in the shell's job group. (Breaking: Nushell integration requires Nushell 0.113+; on Ctrl-C wt dies by the signal instead of exiting 130, so a calling script stops too.) ([#4400
0.80.02026-09-27
Release Notes
Improved
- wt list and the wt switch picker settle faster in repos with many worktrees: rows fill from the top down, and on macOS with core.fsmonitor=true no daemon start is forked for a worktree whose daemon is already running. On a 50-worktree repo the picker's per-row work starts at 47ms, down from 120ms. (#4286, #4290, #4288, #4284)
- CI status skips branches that were never pushed: wt list --full, the picker, and the statusline no longer query the forge for a local branch no remote has, so a repo with 239 local branches made 25 gh calls instead of 249. A PR opened from a branch another clone pushed shows once you fetch. GitLab looks up its project ID once per command, not per row. (#4289, #4216)
- Leaner Codex commit generation: the recommended Codex command moves to gpt-6-luna with reasoning off and unneeded Codex features disabled, so each message sends fewer input tokens. Accepting the first-run Codex prompt now also creates ~/.codex/worktrunk-commit-instructions.txt, which the command reads. Docs (#4261)
- Releases ship a signed packslip: each GitHub release includes packslip.sigstore.json, a manifest of the archives' digests, platforms, completions, and agent skills, signed by the release workflow so installers such as mise can verify downloads before unpacking. (#4301, thanks @jdx)
Fixed
- A squash whose commit fails keeps the branch: when a git commit hook or signing stopped the commit in wt merge or wt step squash, the branch was left at the merge base with its changes staged, recoverable from the reflog. Now the branch is untouched. (Breaking: git commit hooks run on a detached HEAD during a squash.) (#4206, thanks @Bennyjitsu for finding it in #4193)
- wt step diff when local main lags origin/main: for a branch built on the newer origin/main, the diff included the upstream commits in between. It now uses the same base as wt merge. (#4281, thanks @starlightromero, whose #3519 reported the same case for wt merge)
- Submodules with an ignore setting: when submodule.<name>.ignore or diff.ignoreSubmodules hid uncommitted changes inside a submodule, wt remove and wt step prune read the worktree as clean and removed it along with them; they now keep it. With ignore=all, wt step commit and squashes now commit a staged submodule pointer instead of leaving it staged. (#4252, #4192, #4206, #4303, thanks @Duang777)
- Worktrees whose .git was deleted but whose directory remains (macOS $TMPDIR cleanup can do this): wt remove <branch> failed, and wt step prune skipped or errored on them. Both now unregister the worktree and leave its files; one with staged changes or an operation in progress is kept by wt step prune, and refused by wt remove without --force. (#4228, #4226, #4302)
- wt merge --no-squash with a stage mode that stages nothing: with --stage=none and only unstaged changes, --stage=tracked and only untracked files, or only dirty submodule contents, it stopped at "Nothing to commit"; it now merges the branch's commits. (#4251, thanks @Duang777)
- Nushell 0.116: the first wt tab completion in each session printed a Positional completer input deprecated warning. The completer now takes Nushell's new input and still works on older versions. (#4297)
- Hook-log JSON's branch field: wt config state logs --format=json put the sanitized log directory name (feature-x-x2d) there; it now holds feature/x, or null when no local branch owns the directory. (Breaking: filters on the sanitized form need updating.) (#4279)
- wt config show --full outside a git repository: it printed only a git rev-parse failure; it now shows the user config and diagnostics. (#4255)
- Fixes for unusual git setups: --no-ff merges are signed when commit.gpgSign is set without a value or only for one worktree; after git worktree repair, --separate-git-dir repos put worktrees beside the work tree and wt remove works (fixes #4235); removing an orphan worktree before its first commit no longer errors; with a custom merge driver, leaving the picker during a conflict check no longer strands .mergefile files (fixes #4273). (#4206, #4236, #4233, #4274, thanks @zengzheqing and @brndnmtthws for reporting)
- Fixes for unusual names and values: branch and remote names starting with -; non-UTF-8 file names in wt step copy-ignored, promote, push, and merge; a bare ~ in CLAUDECONFIGDIR, or Windows HOME differing from USERPROFILE; post-remove hooks after a pre-remove hook breaks the user config; multiline wt config state vars values, which vars list and {{ vars.* }} truncated and wt config state clear failed on. (#4253, #4267, #4277, #4257, #4259, #4262, #4247, #4276, #4272, thanks @Duang777 and @hiro-nikaitou)
Documentation
- Help and recipe corrections: wt step squash --stage=none and the wt step prune skip list now describe what they do, and the "Database per worktree" and "Monitor hook logs" tips use commands wt accepts. (#4268, #4279)
- The worktrunk skill's sub-Agent example runs hooks: it dropped --no-hooks, so pre-start provisioning runs in agent-spawned worktrees, with guidance on when to add it back. (#4219, thanks @passitalong)
Internal
- Library API changes (Breaking library API): LocalBranch gains pushremote; GitError gains StaleWorktreeHoldsWork and a repairableat field on WorktreeMissing; Repository::capturerefswithaheadbehind is removed; git::parseporcelainz and Repository::changed_files now use bytes. (#4228, #4259, [#4289](https://github.com/max-sixty/workt
0.79.02026-09-21
Release Notes
Improved
- wt switch --no-cd -x <program> runs the program in the worktree: it started in the invoking directory, so a program meant for the new worktree had to be handed {{ worktree_path }}. --no-cd now governs only your shell — wt switch feature --no-cd -x code -- . opens the worktree in an editor and leaves your terminal put. (Breaking: a program that wanted the invoking directory needs an explicit cd.) Fixes #4042. (#4179, thanks @yajo for reporting)
- The picker's preview tabs show which have content: a tab's label is dim when that tab is empty for the selected row, and underlined when it's the active tab. Availability sat on the tab's number alone, so every inactive label read dim whatever the row held. (#4174)
- Messages name the destination, and end with the command to run: without shell integration, wt merge and wt remove printed Cannot change directory — shell integration not installed and no path; the warning now names the worktree, as wt switch already did. "Cannot determine default branch" had three wordings, and three recovery hints put their command first rather than last. (#4163, #4155, thanks @blrain3 for reporting)
- /wt-switch-create no longer stalls at Claude Code's path-entry dialog: Claude Code confirms every entry outside .claude/worktrees/, which is every Worktrunk worktree, so a background session waited for someone to attach. The plugin now approves entry into a worktree at its worktree-path location, and reads a bare name as the repo or the branch. (#4158, #4159)
- wt config update --output accepts the config being migrated: it refused that path when the migration dropped approved-commands. Those approvals now move to approvals.toml first, exactly as the in-place update moves them. (#4204)
Fixed
- wt merge on a dirty worktree describes what it commits: the generated message covered the branch's earlier commits and said nothing about the work the merge had just staged into the same commit. wt step squash --dry-run now previews what the run would commit, as wt step commit --dry-run does. (#4178)
- A config save no longer rewrites the rest of your config: declining a prompt, or setting the commit-generation command, rewrote the whole file, dropping values written at their default and re-spelling hook pipelines and dropping their comments. It also reloaded that file over the in-memory config, so wt step commit --config-set 'commit.stage="none"' staged everything once it had asked about commit generation. (#4160, #4156, #4151)
- Diff display config no longer reaches Worktrunk's prompts or its integration check: color.ui = always and diff.external reached the commit, squash, and summary prompts. Two corner cases also made wt remove and wt step prune delete an unmerged branch: diff.relative set and a Worktrunk command run from a subdirectory the branch's changes sat entirely outside, or submodule.<name>.ignore = all (git config or .gitmodules) and a submodule bump as its only change. (#4146, #4147)
- Auto-staging names a wholly untracked directory once: since 0.78.0 the warning expanded each one into its files, so wt step commit, wt step squash, and wt merge listed 300 paths for a directory of 300 new files. The listing also stops at ten rows, with a hint counting the rest. (#4152, thanks @Duang777)
- Valid empty per-project sections stop warning: [projects."host/org/repo".list] is accepted on load, yet every command reported unknown field projects.host/org/repo.list for it — as did the other six per-project sections. Fixes #4115. (#4123, thanks @zach-hammad-vs for reporting)
- An unstaged rename counts as a change: after git add -N, git reports the rename in the worktree column ( R), which Worktrunk read only in the index column. A worktree whose only change was such a rename showed clean in wt list, and wt merge's overlap check on the destination misparsed the old path. (#4190, #4191, thanks @Duang777)
- wt merge --no-ff and wt step push --no-ff sign their merge commit under commit.gpgSign: they built it with commit-tree, which ignores that setting. wt step rebase and wt merge also pass --no-update-refs, so a rebase rewrites only its own worktree's branch. (Breaking: rebase.updateRefs no longer moves other local branches stacked in the rebased range.) (#4146)
- wt step prune --min-age guards branches with no reflog: they got no age guard at all, so in a bare repository — where core.logAllRefUpdates defaults off — a branch that had just arrived at an integrated commit was pruned on the next run. Age now comes from the mtime of the file holding the ref. (#4154)
- The picker preview renders the whole pane: a fenced code block with no closing fence dropped everything from the fence to the end of a PR description or comment, and the pane rendered one column wider than skim paints, so a wrapped line flush against the right edge lost its last character. (#4178, #4183)
- wt config plugins pi install expands ~ in $PICODINGAGENTDIR: Pi expands it, so PICODINGAGENTDIR="~/x" means a directory under home, where Worktrunk created a literal ~ directory under the current directory. (#4151)
- wt config create no longer replaces a dangling symlink at the config path: it wrote a regular file over the link, detaching whatever owns it. It now fails and names the link. (#4204)
Documentation
- The picker demo shows the CI and summary columns filling in: alt-p widens the rows, the cells land behind them, then it pages a diff, a PR's comment thread, and the PR itself. The homepage omnibus demo opens on the picker. Docs (#4178)
- The docs got a sweep: help text drops internals, deprecated behavior, and edge cases the command reports itself when they happen — wt step diff's "How it works", the wt hook --var paragraph, and the wt config state ci-status page among them. Config samples now name the file they belong in as the block's tab. (#4157, #4153, #4173, #4177)
Internal
- Library API rework (Breaking library API): cargo-semver-checks fails five lints — BranchDiffSpec swaps revs/workingbase for diffbase, UserConfig::save_to gives way to the ConfigEdit mutations, config::UnknownAnalysis is gone, `Repositor
0.78.02026-09-16
Release Notes
Improved
- Hook scripts read renamed context keys: the JSON piped to every hook drops worktree, reporoot, mainworktree, and mainworktreepath in favour of worktreepath, repopath, repo, and primaryworktreepath. Config templates migrate on load, so hooks written with the old names keep working; scripts reading the JSON, --execute templates, and --var overrides need updating. (Breaking.) (#4080)
- wt config plugins pi installs for Pi; oh-my-pi moves to wt config plugins omp: pi install wrote an oh-my-pi hook, which Pi never loads; Pi users re-run it to get a Pi extension. Existing oh-my-pi hooks keep working, listed as outdated until wt config plugins omp install. (Breaking: scripts that run wt config plugins pi for oh-my-pi need omp.) Docs (#4135, thanks @ashebanow)
- Interactive prompts no longer open with a blank line: wt config shell install, wt config plugins claude install, and the commit-message setup offer in wt merge began with one. (#4059)
- wt config show gives a single wt config shell install hint: an outdated wrapper, a fish wrapper at the deprecated conf.d path, and missing fish completions each printed their own hint. The zsh compinit snippet is now syntax-highlighted. (#4059)
- wt config update --output <path> confirms the write: it now prints ✓ Wrote user config migration @ ~/migrated.toml where it printed nothing. --output=- stays silent. (#4053)
Fixed
- Removal no longer discards untracked files hidden by status.showUntrackedFiles = no: under that setting, a worktree holding only untracked files looked clean, so wt remove, wt merge, and wt step prune removed it without --force. Safety checks now always count untracked files, and auto-staging warnings list every file they stage. (#4111, thanks @Duang777)
- wt remove and wt merge check the right worktree under an inherited GITDIR: run from a !wt git alias or a git hook, which pass GITDIR down, wt remove could delete a dirty worktree without --force, and wt merge could refuse a clean merge. Both now check the worktree they act on. (#4082, thanks @Duang777)
- wt merge no longer deletes a worktree locked with git worktree lock: the merge completes and prints Worktree preserved (locked). Every Worktrunk removal refuses a locked worktree, even with --force; run git worktree unlock <path> first. (#4073, thanks @Duang777)
- wt config update no longer overwrites a config edited while its prompt is open: it wrote the pre-prompt snapshot over the newer file and reported success. It now fails with the path and leaves the newer file; re-running previews the current contents. (#4127, thanks @zach-hammad-vs for reporting)
- wt step prune no longer deletes a branch created minutes ago from an older commit: --min-age aged a branch without a worktree by its commit's date, so a new branch off a day-old commit was pruned on the next run. Age now comes from the branch's oldest reflog entry. (#4077)
- A comment or blank line above an inline config section no longer breaks the config: a save or migration that rewrote commit = { … } as [commit] put that line inside the brackets. wt switch, wt merge, wt remove, and several wt step commands then failed with Failed to load config; wt list warned and ran without the config. Saves also dropped unrecognized keys from inline sections. (#4120, thanks @zach-hammad-vs for reporting)
- Config saves keep the trailing comment on a value they rewrite: a save such as answering the commit-generation offer dropped the comment after each value it rewrote, including one after an inline section's closing brace. (#4138)
- Ignored project-config keys warn from every worktree: an unknown key in .config/wt.toml, or one that belongs in user config such as [merge], warned only in the main worktree, so wt merge from a linked worktree ignored it silently. (#4145, thanks @ee-prog for reporting)
- wt merge skips post-commit when it removes that hook's worktree: before, the hook ran in the primary worktree when the removed one sat inside the repository, and otherwise didn't run. The merge now prints ▲ Skipped post-commit; move work that must finish there to pre-remove. (#4049)
- wt switch recovers from a removed worktree in a bare repository: wt switch and the picker surfaced git's raw fatal: Unable to read current working directory instead, and the removed-directory message lacked wt switch ^. (#4067)
- wt config shell install fish puts the wrapper where fish reads it under a custom $XDGCONFIGHOME: the wrapper always went to ~/.config/fish/functions/, so install reported success while wt switch changed no directory. On Windows, the fish completion moves from %APPDATA%\fish\completions to the wrapper's directory. (#4107)
- wt config shell install zsh honours $ZDOTDIR only when it is absolute: an empty or relative value resolved against the current directory, so install appended to a .zshrc under the current directory and reported success while the file zsh reads went untouched. A non-absolute value now falls back to $HOME. (#4085)
- wt config shell install for Nushell no longer deletes a wt.nu under the current directory: an empty or relative $XDGCONFIGHOME resolved against it, so install removed a nushell/vendor/autoload/wt.nu beneath it unread, as a stale wrapper. It now falls back to ~/.config, which also fixes unread tea logins that let wt switch pr:<n> treat a Gitea host as GitHub. (#4104)
- wt switch --create names the branch a failed worktree add leaves behind: git worktree add -b writes the branch ref before populating the worktree, so a failure in between left the branch with no worktree, and the next attempt reported a name collision. The error now says how to delete or reuse it. (#4109, thanks @technicalpickles for reporting)
- wt config shell install no longer tells an already-wrapped shell to restart: reinstalling from a shell with the wrapper loaded, such as after a version bump, printed ↳ Restart shell to activate shell integration. (#4059)
- wt config shell install migrates a deprecated conf.d fish wrapper when ~/.config/fish/functions is missing: before, only wt config shell install fish migrated it. wt config show now reports that wrapper even when fish isn't on PATH. (#4059)
- Plugin status and uninstall checks ask the agent CLIs: wt config show could list an installed Claude Code plugin or Gemini CLI extension as not installed, and wt config plugins claude|codex uninstall could report success over a marketplace still configured. A repeat Claude uninstall now runs both removals instead of stopping at `Plugin not installe
0.77.02026-09-08
Release Notes
Improved
- wt list sizes and aligns its columns to the terminal width: one long branch name no longer sizes the Branch column for every row — it caps at 32 characters and elides, so a narrow terminal keeps its other columns instead of degenerating into a branch list. Remote⇅ stops holding blank space open in a repo with no remote, and alignment now follows the value type. --format=json still carries the whole name. (#3998)
- wt list names what it hid and marks detached worktrees: the summary reads hidden: Path, Commit instead of counting columns, a detached worktree shows ⊘ rather than borrowing ⚑, and a prunable row leaves its cells blank instead of showing loading dots that never resolve. (#3998)
- wt list --format=json gains a marker field and stops reporting detached twice: the branch marker set by wt config state marker is now readable without parsing it out of symbols, and state no longer reports a detached worktree as branchworktreemismatch — the sibling detached boolean carries it. (#3998)
- wt list --format=json now defaults to schema 2: callers get the envelope with repository metadata and orthogonal per-item facts without configuring [list] json-schema. wt list statusline --format=json follows the same key. Set json-schema = 1 to retain the original bare-array format. (Breaking.) (#4038)
- wt config show fails on a broken config: it exits 1 on an unreadable or invalid config, an invalid [list] columns, or an invalid approvals.toml. Unknown keys and deprecations stay warnings and still exit 0. A new APPROVALS section counts project commands awaiting approval. (Breaking: it always exited 0 before.) (#3999)
- Config migration output names its destination: wt config update --output <path> writes the migration artifact to that file instead of applying it in place, and --output=- writes it to stdout. Output mode includes project config when run from a linked worktree. (Breaking: --print was removed.) (#4021)
- Pi joins the agent integrations: wt config plugins pi install writes an activity hook to ~/.omp/agent/hooks/pre/worktrunk.ts, so Pi sessions show 🤖/💬 markers in wt list like Claude Code, Codex, OpenCode, and Gemini. $PICONFIGDIR, $PICODINGAGENTDIR, and named $OMPPROFILE/$PI_PROFILE profiles are honored. Docs (#3594, thanks @adity982, and @ortonomy for the request)
- wt config plugins codex install installs the plugin too: it previously registered the marketplace and left you to run /plugins in Codex yourself. Uninstall follows for both Codex and Claude Code, removing the plugin and then the marketplace, so the two commands are inverses. (Breaking: Codex uninstall previously left an installed plugin alone; Claude uninstall now also removes the marketplace.) (#4019)
- wt step copy-ignored says whether it reflinked or copied: the summary now reads (reflinked, no extra disk), (full copy), or a partial count, so the same 29.5 GB line distinguishes a free copy from one that actually wrote the bytes. --format=json gains reflinked and written. (#4025)
- wt config update removes keys its destination cannot hold: migrating [select] or [commit-generation] drops keys the new section has no field for, reporting each one. Otherwise the key landed at a path the user never typed and warned on every command, with no way to clear it. (#3994)
Fixed
- A failed wt switch pr:<n> against a fork no longer deletes a branch it didn't create: the rollback force-deleted the branch on any setup error, including the one case where the branch wasn't its own — a pre-switch hook or a concurrent session claiming the name in the window after the forge answered. Branch and worktree are now created in one git worktree add -b, which writes nothing when the name is taken. (#3984)
- Codex on Windows: the activity hooks no longer fail on every event: each hook led with a bare bash, which cmd.exe resolves to the WSL launcher rather than Git Bash, so every event raised a Hook failed banner. Hooks now run wt.sh through a shim that finds Git Bash by path. (#4008, fixes #4007, thanks @McNultyyy for reporting and diagnosing)
- Templates leave branch unset in a detached worktree: aliases and hooks rendered {{ branch }} as the literal HEAD, which git resolves, so an unguarded template ran against the wrong ref — the reported case ended in git push origin --delete HEAD, which git refused. base and target follow. (Breaking: an unguarded {{ branch }} now errors.) (#4010, fixes #4009)
- The OpenCode plugin loads under OpenCode 2: OpenCode 2's loader dropped the old bare-function export silently, so activity markers stopped appearing. The plugin now exports one object carrying both the v2 setup and v1 server entry points, and spawns wt through node:child_process rather than the Bun shell. Re-run wt config plugins opencode install to pick it up; the one file covers OpenCode 2 and OpenCode 1.16 or later. (#4018, thanks @pragmaticivan for the request)
- wt step commit --branch builds its prompt from the worktree being committed: prompt-building git plumbing ran in the invoking worktree, so with nothing staged there the generated message described none of the changes it committed. wt step relocate --commit had the same bug, and --dry-run/--show-prompt now preview the worktree --branch names. (#3996)
- LLM commit messages keep diffs for quoted paths: a file whose name git quotes in the diff --git header — any non-ASCII name under the default core.quotePath — had its whole diff section dropped from the prompt once the diff exceeded the 100 KB budget. (#4041)
- An unparsable user config is reported, not a panic: wt config show --format=json and --full, plus wt step prune, relocate, eval, and for-each, panicked on a debug build and printed a bare ✗ Command failed on a release one. They now report ✗ Failed to load config with the parser's caret diagram. (#4002)
- wt switch --execute names the directory the program actually starts in: the Executing (--execute) @ … header rendered the worktree the background hooks run in, not the program's own directory — which differs under --no-cd or a switch from a subdirectory. The path is omitted when the shell already stands there. (#4043, fixes #4042, thanks @yajo for reporting)
- wt config plugins claude|codex uninstall is safe to re-run: removing an already-absent marketplace reported a failure, and a Claude uninstall that removed the plugin and then failed left the marketplace with no way to clear it. Both now finish a half-done uninstall and succeed when there is nothing left to remove. (#4033, [#40
0.76.02026-09-01
Release Notes
Improved
- wt switch --execute takes a program, not a shell string: -x names one program, with everything after -- passed as literal argv. Worktrunk spawns it as a child rather than running it in your interactive shell, so shell functions and its cd no longer reach you. -x sh -- -c '…' recovers shell syntax, not functions. (Breaking: existing -x strings, plus WORKTRUNKDIRECTIVEEXECFILE and WORKTRUNKSHELL.) (#3977, closes #2860, fixes #3944, thanks @omgreenfield for testing the migration path)
- Worktrunk decides tracking for the branches it creates: a new branch gets an upstream only when its name matches the remote branch it starts from, whatever branch.autoSetupMerge says: --create release --base origin/release tracks, --create feature --base origin/release does not. Under autoSetupMerge = false this previously exited 128 after creating the worktree. (#3913, #3950, fixes #3937, thanks @mjakl for reporting and diagnosing)
- Retired config keys are no longer read: commit.generation.template-file and squash-template-file are gone — put the file's contents in template or squash-template. There is no migration: the key stays, warns as unknown on every load, and the prompt reverts to the built-in default. Setting both template and template-file used to fail the load; it now loads template. switch.picker.timeout-ms likewise warns rather than being stripped. (Breaking.) (#3949)
- Project aliases and hooks can use wt switch --execute: their bodies previously refused it. Worktrunk now starts the program itself instead of handing shell text to your shell, and the command-approval gate remains the control on project-defined commands. (#3977)
- Established automation and LLM customization interfaces are now stable: wt step eval, wt step for-each, wt step prune, LLM branch summaries, wt config state vars, and commit.generation.template-append are no longer marked experimental. (#3949)
- wt step prune stages removals concurrently again: only the git worktree remove teardowns serialize, so dirty checks, fsmonitor shutdown, and the trash rename overlap. The prune_e2e/live benchmark median went from 377 ms to 222 ms. (#3954)
- wt config approvals add reads as a review rather than a warning: it opens with a cyan Review 1 command for repo: instead of a yellow ▲ … needs approval to execute. The execution-time gate is unchanged. (#3953)
Fixed
- The wt switch picker survives a preview longer than 65,535 lines: skim keeps the pane's line count in a u16, so a larger diff aborted wt with exit 101 seconds after it painted — on screen the picker looked like it closed by itself. Panes now cap at 60,000 lines and say so. (#3959, fixes #3958, thanks @sandertammesoo for reporting and diagnosing)
- The nushell wrapper propagates exit codes without a POSIX shell: it reported a failing wt by spawning sh, so on Windows every failing command printed a `Command sh not found trace and exit 1 whatever the real code was. Nushell's wrapper is a static file, so rerun wt config shell install to pick this up. (#3945)
- The picker no longer shows one detached worktree's diff for another: previews were keyed by branch name, so every detached row shared the key (detached). Rows now key by canonical identity, which also lets wt remove feature ~/repo.feature plan that worktree once instead of failing on the second. (#3926)
- wt switch pr:<n> gives pre-switch hooks the real branch: branch and target held the literal pr:3933, and targetworktreepath was unset even when that branch had a worktree. prnumber and prurl are now set for same-repo PRs, not only forks. (#3941, fixes #3934, thanks @robsonpeixoto for reporting)
- wt config shell install no longer replaces an rc file created while it was checking: the missing-file path refuses to clobber and fails with a rerun hint. A dangling rc symlink is preserved and rejected rather than overwritten. (#3929)
- wt switch --create X --base pr:<n> works when the PR branch has no local branch: the fetch writes only the remote-tracking ref, and git won't expand a bare name to one, so the command failed with No branch, tag, or commit named …. The base is now named under its remote. (#3951)
- wt switch --create streams git worktree add instead of stalling: where a sandbox denied the timed wait's pipe or sigaction, the command waited out git's entire run and then reported Failed to wait for command for work that had already succeeded. (#3881)
- Claude Code activity markers stay on the session's launch worktree: the plugin's marker hooks pass -C "$CLAUDEPROJECTDIR", so a cd mid-session no longer marks another repository. A session launched outside a repository gets no marker. (#3956)
Documentation
- --execute documents how to get a variable into a shell body: each position is substituted verbatim, so a variable spliced into sh -c text is re-parsed by that shell and splits on spaces. The help now shows passing it as a separate argument and referencing it positionally. (#3977)
- The clone-local default-branch override is documented: wt config state default-branch set writes worktrunk.default-branch to the clone's local git config — machine-local, never committed, and shared by every linked worktree. set warns when the branch isn't checked out locally, and wt list skips its comparisons until it is. (#3947, #3948, fixes #3946, thanks @danielo515 for the request)
- The docs site renders code with the CLI's own colors: terminal examples reproduce each snapshot's exact ANSI roles, while command references and the homepage command comparison get clap and shell syntax color from paired light/dark themes. File excerpts gain a path tab, and code wraps instead of scrolling on mobile. (#3936, #3938, #3939, #3942)
- The pre-start upstream example uses --base: it showed wt switch --create feature origin/feature, which wt switch rejects. (#3955)
Internal
- Library API rework (Breaking library API): cargo-semver-checks fails ten lints — RemoteRefProvider and its provider structs give way to ForgeKind dispatch, the --execute rework retires the fish and PowerShell ShellEscapeMode escapes and WORKTRUNKSHELLENVVAR, BranchRef swaps fullref/wor
0.75.02026-08-27
Release Notes
Improved
- The wt switch picker opens on one unified diff: local rows combine committed, staged, unstaged, and untracked changes. Tab skips empty subsidiary views, while Alt-1 through Alt-8 retain direct access. Docs (#3865)
- Git 2.43 is now the minimum supported version: older Git exits with an upgrade message before Git-dependent commands run; wt config shell remains available so shell startup continues. (Breaking.) (#3895)
Fixed
- wt list no longer grows .git/objects during advisory conflict checks: untracked content stays visible but is excluded from synthetic trees, and all probe-only objects use temporary storage. (#3906, fixes #3883, thanks @srobroek for reporting and the original fix)
- HEAD± counts untracked files without inflating moves: wt list, the picker, and statusline include untracked lines. Tracked deletions paired with untracked destinations count as renames, so pure moves are line-neutral and edited moves show only their edits. HEAD± now always detects renames, regardless of diff.renames. (#3925)
- Timed child processes no longer abort wt in restricted sandboxes: wt switch --create, picker pagers, and other bounded commands survive denied signal-handler wakes. TERM→KILL cleanup also returns promptly once the process group is gone. (#3857, #3887, fixes #3856, thanks @tomascamargo for reporting)
- JSON list output ignores display-column gates: [list] columns no longer makes wt list --format json contact a forge or generate summaries. CI requires --full; summaries also require [list] summary = true and a configured generator. (Breaking: schema 1 loses config-driven ci and summary fields.) (#3812, thanks @emeren for the request)
- Long Windows paths compare consistently: paths beyond 260 characters could retain a \\?\ prefix and appear to be on another drive. copy-ignored then refused them, while switch, remove, merge, and relocate landed at the worktree root instead of the original subdirectory. (#3899, fixes #3898, thanks @Persedes for reporting and verifying the fix)
- Shell configuration rechecks before it writes: overlapping installs lock and reread rc files; Fish completion installs preserve files created after preview; uninstall applies only previewed rc removals and rejects changed Worktrunk-owned files. (#3853, #3924)
- -vv profiles exclude their own collector commands: command counts and cache summaries no longer include duplicate-looking work performed only to assemble the diagnostic report; raw traces still retain it. (#3900)
- Fenced HTML comments survive picker Markdown rendering: PR descriptions and comments now preserve fenced <!-- … --> lines; fenced <!-- wt list … --> markers also no longer affect the following block. (#3908)
Documentation
- Agent CLIs without a plugin can publish activity markers: the integration guide now specifies the session-start, turn-end, and session-end calls, working-directory requirement, error guard, and cleanup contract. Docs (#3848, thanks @AsafMah for requesting generic-agent guidance and @ortonomy for the related Pi use case)
- Agent guidance explains worktree selection: commands that name a branch already select its worktree; -C changes repository context and is needed only for commands without a worktree selector or callers outside the repository. (#3890)
- The wt up recipe safely updates dirty worktrees: it fetches all remotes, fast-forwards dirty branches without autostash, rebases clean branches, and continues past an ordinary refusal or a failed remote. Docs (#3882)
- The docs site has a new responsive design: rebuilt on Astro and Starlight while preserving public routes, anchors, and crawler URLs; generated reference pages remain synchronized. (#3866)
Internal
- Library API rework (Breaking library API): BranchDiffSpec gained working_base, while remote-URL, shell-path, branch-push, approval, temporary-index, and repository helpers were removed. (#3833, #3853, #3865, #3866, #3875)
- Codex loads repository maintainer skills: .agents/skills now exposes the canonical .claude/skills tree; checkouts without symlink support keep the existing limitation. (#3903)
Install worktrunk 0.75.0
Install prebuilt binaries via shell script
``sh
curl --proto '=https' --tlsv1.2 -LsSf https://github.com/max-sixty/worktrunk/releases/download/v0.75.0/worktrunk-installer.sh | sh && wt config shell install
`
Install prebuilt binaries via powershell script
`sh
powershell -ExecutionPolicy Bypass -c "irm https://github.com/max-sixty/worktrunk/releases/download/v0.75.0/worktrunk-installer.ps1 | iex"; git-wt config shell install
`
Install prebuilt binaries via Homebrew
`sh
brew install worktrunk && wt config shell install
`
Download worktrunk 0.75.0
| File | Platform | Checksum |
|--------|----------|----------|
| worktrunk-aarch64-apple-darwin.tar.xz | Apple Silicon macOS | checksum |
| worktrunk-x86_64-apple-darwin.tar.xz | Intel macOS | checksum |
| worktrunk-x86_64-pc-windows-msvc.zip | x64 Windows | checksum |
| worktrunk-aarch64-unknown-linux-musl.tar.xz | ARM64 MUSL Linux | checksum |
| worktrunk-x86_64-unknown-linux-musl.tar.xz | x64 MUSL Linux | checksum |
Install via Cargo
``sh
cargo install worktrunk && wt c
0.74.02026-08-14
Release Notes
Improved
- A JSON bool or null nested in a vars value renders as True, False, or None: minijinja 2.22 adopted Jinja2's spelling for all three, so {{ vars.<key>.<field> }} into an object set by wt config state vars set prints those words. A var whose whole value is true stays a string. (Breaking: the old spellings were true, false, and none.) (#3795)
- wt config approvals add --yes records approvals without a terminal: the command refused every non-interactive run, so a container or CI job could not pre-approve a project it had just cloned. --yes now lists what it trusts and writes it, and a save it cannot make fails the command rather than warning and exiting 0. (#3819)
- remoterepo names the repository as the remote spells it: repo is the directory on disk, so a renamed clone reports the new name. {{ remoterepo }} takes it from the primary remote's URL, available everywhere owner is and unset when no remote parses. (#3745, thanks @canac)
Fixed
- wt remove --force deleted a live worktree moved onto another's registered path: the ownership gate asked which repository the occupant belonged to, which a sibling worktree of this repo satisfies, so the directory went to trash with its uncommitted work. The directory and the registration must now name each other, as git worktree remove requires. Also reaches wt merge, wt step prune, and picker removal, and refuses a bare repository's path that --force previously deleted. (#3808)
- wt step copy-ignored skips a source file that vanishes mid-copy: the walk collects leaves, then copies them in parallel, so a file removed in between — a concurrent build rewriting target/, say — aborted the batch, under --force after deleting the destination. The source is stat'd before anything is removed now. (#3744, fixes #3743, thanks @dataders for reporting and fixing)
- wt step promote refuses to delete a source it did not fully copy: across filesystems the move copies then deletes, and an entry with no copy — a socket or FIFO, or a file a concurrent build rewrote mid-walk — was deleted anyway. The copy now reports what it skipped, and a non-zero count fails the move. (#3820)
- An ssh:// remote's host comes from the authority, not a later path segment: the search for the userinfo @ covered the whole remainder, so ssh://git@attacker.example/owner/repo@github.com/org/repo.git resolved to github.com and borrowed that repository's approvals while git dialed the attacker. Every URL form now splits the authority first, so @ in a namespace parses instead of being rejected. (#3813)
- WORKTRUNK env vars and --config-set outrank a [projects."…"] entry: layer and specificity resolved separately, so a project entry answered for the global key whichever layer set it — WORKTRUNKWORKTREE_PATH could not override a project's worktree-path. Both invocation layers now apply at either scope, and the help text gained a precedence section. (#3790, fixes #3788, thanks @jeremy0dell for reporting)
- Shell completions register under the --cmd name: wt config shell init zsh --cmd wot renamed the wrapper, but clap emitted the registration under wt. Bash and zsh completed nothing and regenerated the script on every TAB, PowerShell registered the wrong command, and zsh's compdef handed worktrunk's completer to the other wt. (#3817, fixes #3816, thanks @sbarre for reporting)
Documentation
- The picker's Alt-x never forces: the keybinding table read Remove selected worktree/branch, which sent a reader looking for a force-remove that isn't there. (#3811, raised in #3809, thanks @josh-burton)
Internal
- Library API rework (Breaking library API): GitError::WorktreePathNotOurs gained an occupantregisteredat field, and WorkingTree::ensurebelongstorepo is now ensureholdsthisworktree. (#3808)
Install worktrunk 0.74.0
Install prebuilt binaries via shell script
``sh
curl --proto '=https' --tlsv1.2 -LsSf https://github.com/max-sixty/worktrunk/releases/download/v0.74.0/worktrunk-installer.sh | sh && wt config shell install
`
Install prebuilt binaries via powershell script
`sh
powershell -ExecutionPolicy Bypass -c "irm https://github.com/max-sixty/worktrunk/releases/download/v0.74.0/worktrunk-installer.ps1 | iex"; git-wt config shell install
`
Install prebuilt binaries via Homebrew
`sh
brew install worktrunk && wt config shell install
`
Download worktrunk 0.74.0
| File | Platform | Checksum |
|--------|----------|----------|
| worktrunk-aarch64-apple-darwin.tar.xz | Apple Silicon macOS | checksum |
| worktrunk-x86_64-apple-darwin.tar.xz | Intel macOS | checksum |
| worktrunk-x86_64-pc-windows-msvc.zip | x64 Windows | checksum |
| worktrunk-aarch64-unknown-linux-musl.tar.xz | ARM64 MUSL Linux | checksum |
| worktrunk-x86_64-unknown-linux-musl.tar.xz | x64 MUSL Linux | checksum |
Install via Cargo
`sh
cargo install worktrunk && wt config shell install
`
Install via Winget (Windows)
`sh
winget install max-sixty.worktrunk && git-wt config shell install
`
Install via AUR (Arch Linux)
`sh
paru worktrunk-bin && wt config shell install
``
0.73.02026-08-12
Release Notes
Fixed
- wt remove refuses a registered path that now holds a different repository: A clone made at a stale registration's path was removed whole, uncommitted work included — and wt routed the user there, since the dirty gate read the occupant's files as this worktree's and offered --force as the cure. Removal now compares the directory's git dir against this repository's; --force waives uncommitted changes, not the check for whose directory it is. (#3785)
- The Nix flake names three systems, dropping Intel macOS: nixpkgs drops x86_64-darwin in 26.11, so the flake would stop evaluating there once flake.lock advances past it. (Breaking: nix build, nix run, and the home-manager module no longer resolve on Intel macOS. Release binaries are unaffected and still ship for it.) (#3776)
- NOCOLOR and a redirected stream reach every surface**: Escapes still landed in progressive wt list, -v diagnostics, clap's error tips, and parts of stderr — the deprecation warning, the wt config update preview, the [y/N] prompt — so a redirected log carried escapes on one line and not the next. Color now resolves in one place. (#3777, #3771)
- A trailing separator no longer hides the branch: Git's ref format forbids a name ending in /, and shell completion produces exactly that spelling whenever a docs directory sits beside the branch, so wt switch docs/ never had a candidate. Selectors are normalized before resolution now — wt remove, merge targets, --base, and the pre-switch hook's target alike. (#3785)
- A path holding no worktree is reported as a path: wt remove ../repo.ghost answered No branch named … and pointed at a listing it could never appear in; wt switch and wt merge offered --create <path>, which git rejects; and wt config state marker set --branch <path> silently stored state keyed by a path. (#3773, thanks @judewang for reporting)
- A failed GitLab project lookup carries glab's own verdict: wt switch mr:<n> on a fork MR answered a 401, a 404, and a network failure with the same Failed to fetch project 456, swallowing the output every other remote_ref failure path forwards. A non-project response body now says so too. (#3799)
- A deleted-and-recreated worktree directory is reported, not leaked as git's exit 128: An rm -rf followed by a mkdir passes the Path::exists() probe, so wt switch, wt merge, wt step push, and wt remove walked into a raw git failure. All four now name the missing worktree and the git worktree prune that clears it. (#3785)
Internal
- Library API rework (Breaking library API): cargo-semver-checks fails four lints — GitError gained WorktreeNotFoundAtPath and WorktreePathNotOurs, shifting seven later discriminants, its DetachedHead variant gained a worktree field, ResolvedWorktree gained NoWorktreeAtPath, and Repository::resolveworktreename was removed. (#3785, #3773)
- Tests and benches spawn a pinned wt binary: a concurrent cargo removes and recreates target/debug/wt as it uplifts, so spawns hit a one-off NotFound; every spawn now routes through a hardlinked pin. (#3784, #3792)
- The nix devShell and task setup-web install what the test suite drives: both were missing nushell, pwsh, and jq, which --features shell-integration-tests shells out to. (#3768, #3776)
- Benchmark fixtures reduced to two provenance-based bases: Generated builds a repository locally and Imported copies the pinned rust-lang/rust corpus, with worktree, branch, and remote-ref populations as parameters. (#3761)
Install worktrunk 0.73.0
Install prebuilt binaries via shell script
``sh
curl --proto '=https' --tlsv1.2 -LsSf https://github.com/max-sixty/worktrunk/releases/download/v0.73.0/worktrunk-installer.sh | sh && wt config shell install
`
Install prebuilt binaries via powershell script
`sh
powershell -ExecutionPolicy Bypass -c "irm https://github.com/max-sixty/worktrunk/releases/download/v0.73.0/worktrunk-installer.ps1 | iex"; git-wt config shell install
`
Install prebuilt binaries via Homebrew
`sh
brew install worktrunk && wt config shell install
`
Download worktrunk 0.73.0
| File | Platform | Checksum |
|--------|----------|----------|
| worktrunk-aarch64-apple-darwin.tar.xz | Apple Silicon macOS | checksum |
| worktrunk-x86_64-apple-darwin.tar.xz | Intel macOS | checksum |
| worktrunk-x86_64-pc-windows-msvc.zip | x64 Windows | checksum |
| worktrunk-aarch64-unknown-linux-musl.tar.xz | ARM64 MUSL Linux | checksum |
| worktrunk-x86_64-unknown-linux-musl.tar.xz | x64 MUSL Linux | checksum |
Install via Cargo
`sh
cargo install worktrunk && wt config shell install
`
Install via Winget (Windows)
`sh
winget install max-sixty.worktrunk && git-wt config shell install
`
Install via AUR (Arch Linux)
`sh
paru worktrunk-bin && wt config shell install
``