Get started
Install kimün and measure your first project
One binary called km, with no dependencies and no configuration.
Install
With Cargo, from crates.io:
cargo install kimun With Homebrew, on macOS and Linux:
brew install lnds/kimun/kimun There are also ready-made binaries for macOS, Linux and Windows on the releases page. See the releases →
The first measurement
Go to the folder of your project and ask for the grade. Kimün respects your .gitignore and needs no git repository for static metrics.
$ km score
Project Score: A (91.0)
Files Analyzed: 127
Total LOC: 18,594
Dimension Weight Score Grade
Cognitive Complexity 30% 83.6 B+
Duplication 20% 97.3 A++
Indentation Complexity 15% 93.5 A+
Halstead Effort 20% 87.4 A-
File Size 15% 99.6 A++
Files Needing Attention
57.9 D- src/deps/analyzer.rs Cognitive: 36
63.9 D+ src/cogcom/analyzer.rs Cognitive: 35
72.3 C src/smells/rules.rs Cognitive: 20real output: kimün measuring itself
What to look at next
km hotspots- Where to improve first: complex files that change often.
km impact --since-ref main- What your branch reaches before you merge it.
km score --trend main --fail-if-worse --gate-scope changed- The gate for your continuous integration.
km ai skill claude- The skill that lets your agent use km.
Configuration
None is needed, but km init gives you a .kimun.toml calibrated for your project, where you can tune thresholds and exclusions.
km init Full documentation
Every command, with its options and examples, is in the README of the repository.