Skip to main content

Quickstart

Get zuit running, scan a project, and learn to read the output — in about five minutes. This is the single page to send a new teammate.

Install​

Pick the package manager you already have. All three paths give you the same zuit binary.

Rust — cargo (canonical; ships the binary directly):

cargo install --locked zuit

You'll need cargo installed first. If you don't have it, get it from rustup.rs — the installer sets up cargo and the Rust toolchain in one step.

Python — pip (wheels bundle the precompiled binary, so zuit analyze . works right after install):

pip install zuit

Node.js — npm (installs a launcher; pair it with cargo install zuit or set ZUIT_BIN to point at an existing binary):

npm install -g zuit

Verify the binary is on your PATH:

zuit --version

The pip and npm packages are launchers — they resolve a zuit binary via ZUIT_BIN, a bundled binary, or your OS PATH. The pip wheel already includes a binary; the npm package does not, so install via cargo or point ZUIT_BIN at an existing binary.

Build from source (for the latest unreleased changes):

git clone https://github.com/shubhamkaushal765/zuit
cd zuit
cargo build --release -p zuit-cli

The binary lands at target/release/zuit. Add that directory to your PATH or invoke it directly.

GitHub Action — to run zuit in CI without managing a Rust install, see GitHub Action.

Language support​

LanguageStatusExtensions
RustFull.rs
PythonFull.py
JavaScript / TypeScriptFull.js, .mjs, .cjs, .jsx, .ts, .mts, .cts, .tsx
GoStubNot yet supported

Files in unsupported languages are skipped silently.

Run your first scan​

Point zuit analyze at any directory. No configuration file is required.

zuit analyze ./src

zuit scans every Rust, Python, and JavaScript/TypeScript file it finds, respecting .gitignore and skipping common vendor directories automatically. Results appear in the terminal as soon as the scan finishes.

For a one-off scan that doesn't get saved to history, add --no-save:

zuit analyze ./src --no-save

Read the output​

The terminal report lists findings grouped by dimension and severity. Each finding takes two lines:

[MAINT001-cyclomatic] medium  src/lib.rs:42:1
  function `process_request` has cyclomatic complexity 14 (threshold 10)

The rule ID (MAINT001-...) is stable across releases and used in config and baselines. The severity is one of info, low, medium, high, or critical. Findings are sorted by file, then line, then rule ID — two scans of unchanged source produce identical output.

After the findings, a scoreboard shows one 0–100 score and an A–F grade per dimension:

Maintainability  87.4  B
Security         98.1  A
Complexity      100.0  A
Documentation    73.5  C
TestSmell        91.0  A

For the scoring formula see Severity and scoring. For the full terminal format reference see Terminal output.

See the dashboard​

After any scan, open the interactive history dashboard:

zuit show

This starts a local HTTP server and opens a browser tab with Overview, Scans, Findings, Trends, Diff, Heatmap, and Config tabs. Every zuit analyze run is saved automatically, so the Trends and Diff views fill in over time. See Track trends across releases for the full flow.

Set up a config (optional)​

Once you know which rules and thresholds matter for your project, generate a zuit.toml starter file:

zuit init

This writes zuit.toml to the current directory with commented sections for general settings, per-dimension thresholds, and per-rule overrides. See the zuit.toml reference for every available field.

What's next​