1. Install the CLI
The runner ships as a CLI that runs the same on your machine, in CI and in the cloud.
Install
npm install -g @stakari/cli
stakari --versionTest your localhost in a few minutes
Install the CLI, point it at a target on your machine, approve the plan, and read the findings. No test scripts to write.
2. Log in and create a project
Log in, then create a project for the app you want to test.
A project holds targets, environments, roles, flows, scenarios and runs.
Set up a project
stakari login
stakari init3. Test a target on localhost
Local mode tests a target on localhost. Results are written locally and uploaded only when you choose to.
Describe what to test in plain English. Stakari turns it into a plan you approve.
Run against localhost
stakari run --env local --module web \
"test the login page at http://localhost:3000"4. Bring a model key or use the gateway
In local mode you can supply your own model key or use the Stakari gateway.
Credentials are never logged and never sent to a model as content.
Use your own key
export STAKARI_MODEL_KEY=sk-...
stakari run --env local --module api5. Approve the plan
No scenario runs until you approve the plan. You can remove or edit scenarios first.
The plan shows the estimated credits as a range and the risk classes involved.
6. Fail the build on findings
In CI, run against a dev or staging URL and fail the job when findings at or above a severity you choose exist.
The CLI accepts --env, --module, --scenario, --fail-on and --json. Exit codes are stable and documented.
In CI
stakari run --env staging --module api --fail-on high --jsonCommand names follow the CLI in the SRS. The install command and package name are provisional and will be confirmed at release.