Saved copy of the Lotus Behavioral App Script Health dashboard · snapshot 2026-10-27, scoring version 1 (leaves out no tables) · works offline

Lotus Behavioral App Script Health

Most tables are small and clean, but — tables graded D hold — of the live code. — scores — and is fed by — other tables, so it is the place to start.

Health score, weighted by code size

— / 100

Simple average across tables: —

Live script code

—

lines in — enabled scripts

Functions that are never called

—

lines inside live scripts

Disabled scripts

—

lines across — scripts

Repeated lookups

—

same record loaded again in one script

Copied functions that have drifted

—

of — functions copied into 2+ scripts

Since the last snapshot

Compares this snapshot with the one before it, within one scoring version.

Measure	Previous	This snapshot	Change

Health score over time

Each run of tvsnap health after a snapshot adds a point. Scores compare only within one scoring version.

Scores by table

Sorted by review priority: points lost × size × how many other tables feed it. Click a row to map it below.

Dependency map

Left: tables whose scripts load or create records in the selected table. Right: tables the selected table's scripts load or create. Dot color is each table's grade.

creates records loads only

Selected table

—

Where it loses points

Quick wins

Low-risk cleanups that cut code without changing behavior. Confirm each one in TrackVia before removing anything.

Functions never called

Defined in a live script, with no call to them anywhere in that script.

Function	Script	Table	Lines

Copies that have drifted

Same function name in several scripts, with different code in each copy.

Function	Scripts	Versions	Tables

Lookups loaded more than once

A record fetched by a typed-in name, again, in the same script. Load it once into a variable.

Looked up	Extra loads	Scripts

How the score works

Each table starts at 100 and loses points for what its enabled scripts contain. This is read from the code in the 2026-10-27 snapshot, not from running it, so it measures how hard the code is to maintain and how much it asks TrackVia to load, not run time.

Script size, up to 25: biggest live script over 300, 700, 1,500 or 3,000 lines.
Long function, up to 10: longest function over 100, 200 or 400 lines.
Repeated lookups, up to 20: 2 per record loaded again by the same typed-in name in one script.
Unused code, up to 15: commented-out code plus never-called functions, as a share of live lines.
Disabled scripts, up to 10: 1 per 100 lines kept in disabled scripts.
Copied functions, up to 15: 3 per copy that differs from its twins. Identical copies are free, since scripts cannot share a function.
Hand-built JSON, up to 5: strings that build JSON by hand instead of groovy.json.JsonBuilder.

Grades: A 85+, B 70–84, C 55–69, D below 55. Queries inside loops were also checked and none were found. No tables are left out of the score. Copied functions are counted across live scripts only. tvsnap health computes everything on this page.