fix(completions): update zsh completions for usage v3.1.0 by jdx · Pull Request #8715 · jdx/mise

jdx · 2026-03-22T18:34:57Z

Summary

Usage v3.1.0 changed zsh complete-word output from quoted ('val') to unquoted (val) format and switched the generated completion template from _arguments to _describe -S ''
Updates the prerendered completions/_mise to use the new _describe pattern
Updates e2e/tasks/test_task_completion expected output to match

Test plan

e2e tasks/test_task_completion passes with usage v3.1.0
Zsh completions work correctly with the new _describe approach

🤖 Generated with Claude Code

Note

Low Risk
Low risk: updates only shell completion generation/formatting and the associated e2e assertion; impact is limited to zsh completions output/behavior.

Overview
Updates the generated zsh completion script completions/_mise to match usage complete-word output in usage-cli v3.1.0 by switching from _arguments to _describe and reading completion lines into an array.

Adjusts the e2e/tasks/test_task_completion expected output to reflect the new unquoted completion formatting, and bumps cargo:usage-cli in mise.lock from 2.18.2 to 3.1.0.

^{Written by Cursor Bugbot for commit 508afbe. This will update automatically on new commits. Configure here.}

Usage v3.1.0 changed zsh complete-word output from quoted ('val') to unquoted (val) format and switched the generated completion template from _arguments to _describe. Update the prerendered zsh completion script and e2e test to match. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

gemini-code-assist · 2026-03-22T18:35:09Z

Summary of Changes

Hello, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request updates the Zsh completion logic and associated end-to-end tests to accommodate breaking changes introduced in usage tool version 3.1.0. The changes primarily involve adapting to a new unquoted output format for completion words and a revised completion template, ensuring mise's Zsh completions continue to function correctly.

Highlights

Zsh Completion Output Format Change: The usage tool v3.1.0 now outputs complete-word results without quotes, changing from 'val' to val.
Zsh Completion Template Update: The usage tool v3.1.0 has switched its generated completion template from _arguments to _describe -S ''.
Completion Script Adaptation: The completions/_mise script has been updated to adopt the new _describe pattern for Zsh completions.
E2E Test Adjustment: The expected output in the e2e/tasks/test_task_completion test has been modified to reflect the new unquoted format from usage v3.1.0.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for GitHub and other Google products, sign up here.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

greptile-apps · 2026-03-22T18:36:01Z

Greptile Summary

This PR adapts the Zsh completion wrapper and its e2e test to usage-cli v3.1.0, which changed complete-word output from single-quoted space-separated tokens to bare unquoted newline-separated entries and switched the recommended completion pattern from _arguments to _describe.

completions/_mise: Replaces the old _arguments "*: :(($(command usage complete-word ...)))" one-liner — which collapsed newlines to spaces via $() substitution — with a while IFS= read -r line loop that correctly populates a zsh array, then passes it to _describe 'completions' completions -S ''.
e2e/tasks/test_task_completion: Updates the expected assertion output from quoted ('mise.toml', 'mise.usage.kdl') to bare unquoted values to match the v3.1.0 format.
mise.lock: Bumps cargo:usage-cli from 2.18.2 to 3.1.0.
The initial commit's fix was inadvertently reverted by autofix.ci; a follow-up commit (508afbe) correctly re-applied the _describe change alongside the lock bump, so the current HEAD is in the intended state.

Confidence Score: 5/5

Safe to merge — minimal, well-targeted changes with no logic risk.
All three files are updated consistently: the completion script correctly uses the loop + _describe pattern, the e2e assertion matches the new unquoted output format, and the lock file pins the required usage-cli version. The previous concern about the autofix.ci revert has been fully resolved by the final commit.
No files require special attention.

Important Files Changed

Filename	Overview
completions/_mise	Replaces the old `_arguments` + `$()` inline-expansion with a `while read` loop populating an array fed to `_describe -S ''`, correctly adapting to usage v3.1.0's unquoted newline-per-entry output format.
e2e/tasks/test_task_completion	Updates expected assertion output from single-quoted values (`'mise.toml'`) to bare unquoted values (`mise.toml`) to match the new usage v3.1.0 output format.
mise.lock	Bumps `cargo:usage-cli` from `2.18.2` to `3.1.0`, the version that introduced unquoted newline-separated `complete-word` output.

Sequence Diagram

sequenceDiagram
    participant Zsh as Zsh Completion Engine
    participant Wrapper as _mise wrapper
    participant Cache as Spec Cache File
    participant UsageCLI as usage complete-word v3.1.0
    participant Describe as _describe

    Zsh->>Wrapper: trigger completion for mise
    Wrapper->>Cache: check spec file exists?
    alt spec file missing
        Wrapper->>Cache: mise usage store spec
    end
    Wrapper->>UsageCLI: complete-word --shell zsh -f spec_file
    UsageCLI-->>Wrapper: unquoted newline-separated completions
    Note over Wrapper: read each line into completions array
    Wrapper->>Describe: _describe completions array -S ''
    Describe-->>Zsh: present completion candidates

_{Reviews (3): Last reviewed commit: "fix(completions): bump usage-cli to 3.1...." | Re-trigger Greptile}

gemini-code-assist

Code Review

This pull request updates the zsh completion script to support changes in usage v3.1.0, which now produces unquoted completion values and uses a different completion function. The changes in the completion script and the corresponding test update are correct. I've provided one suggestion to make the zsh script more idiomatic and concise.

gemini-code-assist · 2026-03-22T18:36:55Z

+  local -a completions=()
+  while IFS= read -r line; do
+    completions+=("$line")
+  done < <(command usage complete-word --shell zsh -f "$spec_file" -- "${words[@]}")


While the while read loop is correct, you can use a more idiomatic and concise zsh feature to populate the completions array. The (f) parameter expansion flag is designed for splitting command output by newlines, which simplifies this block of code significantly.

local -a completions=("${(@f)$(command usage complete-word --shell zsh -f "$spec_file" -- "${words[@]}")}")

greptile-apps · 2026-03-22T18:42:09Z

+assert "mise exec -- usage complete-word --shell zsh -f ./mise.usage.kdl -- mise tasks run build -- -c ''" "mise.toml
+mise.usage.kdl"


completions/_mise fix reverted by autofix.ci

The autofix.ci bot (commit a087196) reverted the completions/_mise change that was introduced in the first commit (dd0e2ad). As a result, the net diff across this entire PR leaves completions/_mise identical to the base branch — still using the old _arguments inline-expansion pattern:

_arguments "*: :(($(command usage complete-word --shell zsh -f "$spec_file" -- "${words[@]}" )))"

This contradicts the PR's stated goal. The intended fix — replacing _arguments with a loop + _describe -S '' — was never shipped:

local -a completions=() while IFS= read -r line; do completions+=("$line") done < <(command usage complete-word --shell zsh -f "$spec_file" -- "${words[@]}") _describe 'completions' completions -S ''

With usage v3.1.0 outputting one entry per line (unquoted), the old _arguments pattern uses $() substitution, which collapses newlines into spaces. For completions that include spaces in their values or colon-separated value:description pairs, this will silently mangle the output. The e2e test here passes because mise.toml and mise.usage.kdl happen to be single tokens without spaces, so the broken path is not exercised.

The autofix.ci bot appears to have regenerated completions/_mise using an older installed version of usage (pre-v3.1.0), effectively undoing the manual fix. You likely need to prevent autofix.ci from regenerating this file, or re-apply the _describe change after the bot runs.

The autofix.ci bot reverted the _describe completion fix because mise.lock still pinned usage-cli at 2.18.2. Bump the lock to 3.1.0 so the generated completions use the new _describe pattern. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

github-actions · 2026-03-22T19:14:13Z

Hyperfine Performance

`mise x -- echo`

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
`mise-2026.3.10 x -- echo`	24.5 ± 0.5	23.7	30.9	1.00
`mise x -- echo`	25.1 ± 0.9	23.8	31.1	1.03 ± 0.04

`mise env`

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
`mise-2026.3.10 env`	24.2 ± 0.4	23.3	26.5	1.00
`mise env`	24.3 ± 1.9	23.2	64.5	1.00 ± 0.08

`mise hook-env`

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
`mise-2026.3.10 hook-env`	24.7 ± 0.4	23.9	26.6	1.00
`mise hook-env`	24.8 ± 0.6	23.6	30.5	1.00 ± 0.03

`mise ls`

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
`mise-2026.3.10 ls`	24.5 ± 0.4	23.7	27.4	1.00
`mise ls`	24.8 ± 0.4	23.7	26.3	1.01 ± 0.02

`xtasks/test/perf`

Command	mise-2026.3.10	mise	Variance
install (cached)	155ms	156ms	+0%
ls (cached)	85ms	86ms	-1%
bin-paths (cached)	88ms	88ms	+0%
task-ls (cached)	855ms	823ms	+3%

@jdx

### 🐛 Bug Fixes - **(completions)** update zsh completions for usage v3.1.0 by @jdx in [#8715](#8715) ### Security - **(lock)** block github tool upgrade when provenance is lost by @jdx in [#8706](#8706)

iamkroot · 2026-03-23T17:11:08Z

Did this actually fix anything?
I'm on v2026.3.13 and getting the following on trying to complete tasks-

$ mise run <TAB>
(eval):37: parse error near `<'

naoki-mizuno · 2026-04-20T08:28:55Z

I’m also getting a similar error when I run a completion after mise completion zsh > ~/.local/share/zsh/completions/_mise:

$ mise <TAB>
(eval):73: unmatched '

Name	Version
macOS	26.3.1
zsh	5.9
mise	2026.4.18
usage	3.2.0

I don’t get this error when running eval "$( mise completion zsh )"... not sure what the difference is.

jdx enabled auto-merge (squash) March 22, 2026 18:35

jdx disabled auto-merge March 22, 2026 18:35

gemini-code-assist Bot reviewed Mar 22, 2026

View reviewed changes

[autofix.ci] apply automated fixes

a087196

greptile-apps Bot reviewed Mar 22, 2026

View reviewed changes

jdx merged commit a7e280b into main Mar 22, 2026
25 checks passed

jdx deleted the fix/usage-3.1-zsh-completions branch March 22, 2026 18:59

mise-en-dev mentioned this pull request Mar 22, 2026

chore: release 2026.3.12 #8717

Merged

BrewTestBot mentioned this pull request Mar 22, 2026

mise 2026.3.12 Homebrew/homebrew-core#273674

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

fix(completions): update zsh completions for usage v3.1.0#8715

fix(completions): update zsh completions for usage v3.1.0#8715
jdx merged 3 commits intomainfrom
fix/usage-3.1-zsh-completions

jdx commented Mar 22, 2026 •

edited by cursor Bot

Loading

Uh oh!

gemini-code-assist Bot commented Mar 22, 2026

Uh oh!

greptile-apps Bot commented Mar 22, 2026 •

edited

Loading

Uh oh!

gemini-code-assist Bot left a comment

Uh oh!

gemini-code-assist Bot Mar 22, 2026

Uh oh!

greptile-apps Bot Mar 22, 2026

Uh oh!

Uh oh!

github-actions Bot commented Mar 22, 2026

Uh oh!

iamkroot commented Mar 23, 2026

Uh oh!

naoki-mizuno commented Apr 20, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

		assert "mise exec -- usage complete-word --shell zsh -f ./mise.usage.kdl -- mise tasks run build -- -c ''" "mise.toml
		mise.usage.kdl"

Uh oh!

Conversation

jdx commented Mar 22, 2026 • edited by cursor Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Test plan

Uh oh!

gemini-code-assist Bot commented Mar 22, 2026

Summary of Changes

Highlights

Footnotes

Uh oh!

greptile-apps Bot commented Mar 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Greptile Summary

Confidence Score: 5/5

Important Files Changed

Sequence Diagram

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist Bot Mar 22, 2026

Choose a reason for hiding this comment

Uh oh!

greptile-apps Bot Mar 22, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actions Bot commented Mar 22, 2026

Hyperfine Performance

mise x -- echo

mise env

mise hook-env

mise ls

xtasks/test/perf

Uh oh!

iamkroot commented Mar 23, 2026

Uh oh!

naoki-mizuno commented Apr 20, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

jdx commented Mar 22, 2026 •

edited by cursor Bot

Loading

greptile-apps Bot commented Mar 22, 2026 •

edited

Loading

`mise x -- echo`

`mise env`

`mise hook-env`

`mise ls`

`xtasks/test/perf`