Indicator dependencies

Definitions and studies can import indicators as dependencies — your script reads their series at each bar instead of reimplementing the math.

Why import

• One VWAP / volume / structure indicator, many definitions

• Fix the indicator once, update dependents deliberately

• Cleaner Definition scripts focused on setup logic

Add a dependency

In the Definition Builder or study builder:

1. Open Imports or Dependencies

2. Choose an indicator from Mine (or Community if licensed)

3. Pin a version when the UI requires it

4. Reference outputs in code — e.g. deps.MyVWAP.value

Exact accessor names appear in builder docs and autocomplete.

Empty output

At runtime a dependency may return empty for a bar — no data yet, wrong session, warmup period.

Your Definition should treat “no value” as no emit, not a crash. Studies should exclude or bucket those bars.

Version changes

Bump an indicator version → review dependents. Recollect datasets if emit logic changed.

Conductor may nudge you to recollect after dependency updates.

What good looks like

Shared session-level indicator imported by three definitions. Each definition only encodes what makes its setup unique.

Important distinctions

Dependency indicator = read series inside scripts.

Chart indicator = drawn on Terminal for visual trading — same object, two uses.

Event payload = fields you attach at emit — separate from indicator plots.

Next

1. Build an indicator — create the dependency

2. Definition Builder — import and emit

3. Versions and snapshots — freeze before collect