Glossary
Every Drizz term in one place — DT, Memory, modules, datasets, lockable pools, MAP_ACTION, Vision AI and the rest.
The vocabulary used across these docs and in the product UI, alphabetically.
Covers
Terms used throughout the documentation and in the product UI
Naming
Chat is also referred to as Fathom in older material
Watch out
"Memory" names two different things. Check which one a page means
AUT
App under test — the app Drizz is driving
Blocker rule
A Memory rule that dismisses a known popup automatically. One rule carries one action
Cached step
A step Drizz has resolved on this screen before and replays without calling Vision AI. Lowest-cost step category
Chat
The panel that generates a test from a plain-English description. Referred to as Fathom in older material
Dataset
Named values bound to a test plan at run time and referenced in a script as {{var}}. Nesting is limited to two levels
Desktop app
The Drizz application installed locally to author tests and run them on a connected device
Drizz device cloud
The hosted devices a test plan runs on. One of three run targets, with local and private devices
DT
Drizz Token — the unit Drizz meters usage in, deducted per step. Current rates are on your organization's billing page
execution_id
The identifier returned when a test plan is triggered over the API. Confirms a run started. Not pollable
Lockable pool
A dataset whose rows are leased exclusively to one run at a time. A run waits when the pool is fully leased
MAP_ACTION
A grid- and coordinate-based gesture, for elements Vision AI cannot target normally. Pinch and zoom are limited to specific Android emulator types
Memory
Org-level app context and blocker rules, applied before a run starts. Unrelated to Store and SET variables
Module
A reusable block of steps, defined once and invoked with CALL. Takes parameters, passed by value
Private device
A device reserved for a single organization
Registered app
An app binary uploaded to Drizz, available to test plans and to Memory. Capped at 500 MB
Self-healing
Automatic repair of a failed step, a limited number of times per run. Healed steps are badged in the report
SET
Assigns a value to a variable inside a script
Store
Captures a value from the current screen into a variable
Test plan
A set of tests plus how and where they run — devices, concurrency, app version, dataset binding
Variable
A named value used inside a script. Produced by Store, SET, dataset values or placeholders
Vision AI
The layer that reads the screen and locates elements without selectors or an SDK
Wallet
An organization's DT balance, shared org-wide. States are Healthy, Low and Blocked
Terms with two names
Chat
Fathom
Write Chat. Fathom in older material means Chat
Memory
—
Memory is the org-level app context and blocker rules only. Runtime values from Store and SET are variables
Next
Command index — every command, one table
Known limitations — what Drizz doesn't do
Which variable do I use? — Store vs SET vs dataset vs placeholder
Last updated: 6 August 2026
Last updated
Was this helpful?