kalku run

Measures the files you name: it casts wekufe over them, one at a time, and reports the ones no test caught.

kalku run [files...] [options]

kalku run lib/thing.ex                    # one file
kalku run lib/green.ex --limit 6 --verbose
kalku run --since origin/main             # only the lines the change touched
kalku run --all --shard 1/4               # the whole project, in parts
kalku run lib/thing.ex --format agent     # survivors as NDJSON

Options

Option	What it does
--limit <n>	Cast at most n wekufe: 100 by default, 0 for all of them. A run with --since casts the whole change unless told otherwise
--since <ref>	Measure only the lines changed since that git ref, and block when a wekufe survives on one of them
--all	Measure every source file of the project, all of its wekufe
--shard <i/n>	Cast only part i of n: the same part on every machine, so n jobs cover a project between them
--store-baseline	Store this run’s score as the baseline ratchet = true holds later runs of the whole project to
--format <name>	human (the default), agent, json, or github
-v, --verbose	Say what the run is doing while it does it

Exit codes

Code	Meaning
0	Measured; nothing to report
1	Measured, and your tests have holes
2	kalku could not measure

In CI explains each one.

What to read next

• What a run tells you: how to read the report, and the formats.
• Outcomes and score: what each wekufe can come to.
• In CI: --since, --all, --shard and the ratchet.