Skip to content

Test files

Tests and flows are *.sis.yaml files anywhere under the project. A file with name: and steps: is a test; a file with flow: is a flow. JSON Schema: schema/test.schema.json.

sisyphus: 1 # format version, always 1
name: Add a todo # what should happen; ${vars} allowed (data rows)
app: web # an app from sisyphus.yaml, or an inline app spec
tags: [smoke]
os: [macos] # only on these systems; skipped elsewhere
timeout: 8s # step timeout for this test
step_delay: 1s # pause before each step
retries: 2 # rerun when it fails; a pass on a rerun is "flaky"
a11y: strict # warn | strict | off
vars: { item: Milk }
app_setup: false # skip the apps' own setup/teardown
data: [ ... ] # or a CSV file name: run once per row
setup: [ ... ] # before the steps
steps: [ ... ]
teardown: [ ... ] # after, always
Key Default Notes
sisyphus Required, 1
name Required
app The app the steps work in. It launches on the first step that needs it and closes at the end
tags [] For --tag, --skip-tag and suites
os all macos, windows, linux: the systems the test runs on. Elsewhere it’s reported as skipped, not failed. For a few steps that differ, use if: { equals: ["${os}", macos] } instead
timeout defaults.timeout (5s) How long each step waits for its target
step_delay none Pause before each step
retries defaults.retries (0) --retries and a suite’s retries win
a11y defaults.a11y (warn) See Accessibility checks
vars Test variables; override the project’s
app_setup true false skips every app’s setup/teardown
data A list of rows, or a .csv file next to the test
setup, steps, teardown Lists of steps
sisyphus: 1
flow: login # the name tests call
params: [email, password]
vars: { email: demo@example.com } # defaults for params
steps: [ ... ]

Called with call: login and with: { … }. See Reusing steps with flows.

data:
- { case: empty email, email: "", error: Email is required }
- { case: bad password, email: a@b.c, error: Wrong password }
# or
data: logins.csv

Each row runs as its own test; row values are variables (over vars:, under --var). In CSV, the first line is the header, "quoted, fields" may hold commas and "" is a quote. file#N runs row N (from 1).

Made by Bussin Bytes© 2026 Bussin Bytes Ltd · sisyphus.bussinbytes.io