Getting started
Save a small app, establish a historical baseline, and verify its first running session.
Start with one instrument and one source of order decisions. SJS is the main algorithm path; use a small strategy you can explain from input to requested action.
Create the saved configuration
Create an app in your workspace and choose the intended app type. Select an exact instrument and data interval. Give the app a name that distinguishes it from copies or alternative source revisions.
Use the SJS getting-started strategy for a complete historical baseline, or the market-data probe when you only want to observe incoming bars. The probe does not request orders.
Save the source and settings. An unsaved editor change is not a reliable description of what a new server will load.
Check a historical run
Choose a short window with available data and sufficient initialization history. Inspect chart data, logs, and any trades. Keep the saved configuration and run identity together.
A completed test with no trades can be valid if its conditions never occur. Verify the expected decisions rather than using completion or positive P&L alone as the acceptance check.
Launch the intended environment
Open Run an app and review the selected app, launch target, broker/account if required, and warmup setting. Start with a simulated runtime when validating the running-session workflow.
Select Run App after those choices match the intended environment. Publishing or saving the app does not perform this step automatically.
Verify startup
Open the resulting server and check its status, current data timestamps, and logs. Confirm the instrument and execution environment again. A RUNNING status is only one check; data may be stale and requests may still be rejected.
For the probe, expect diagnostic observations and no orders. For a strategy, trace one decision through request and outcome before drawing conclusions from totals.
Continue with configuration and launching.