Skip to main content
Run the test script for every selected package and module that declares one. The command is standalone and keeps its own cache for test runs.

​
Usage

​
Examples

Test every package and module that has a test script:
Test only selected modules and packages:
Show failed task output in the report:
Force every selected test task to run again:
Run tests from another workspace directory:

​
Options

​
Execution

test discovers workspace targets, applies the package and module filters, sorts the selected targets, and builds a task group from targets that declare test. Targets without a test script are left out. If no selected target declares the script, the command prints Skipped test (no such script) and exits successfully. For a target with a package.json, the task runs bun run test. For a recognized Rust or Python target without a package.json, it runs that target’s default test command directly. A test task whose tests directory has no files is marked skipped.
workspace:run --commands=test delegates to this standalone command, so the cache and behavior match talos test.

​
Caching

Test entries live in var/cache/test. test uses its own fingerprint cache, separate from build, fmt, lint, and workspace runs. Pass --no-cache to bypass both reads and writes.

​
Output

The command prints the task count before running, then a Test report with one row per task that ran or replayed from cache. Cached tasks are marked cached. Failed tasks appear under Failing tasks; use --logs to include their captured output.