Blue Prism Pro Developer key terms

The top-level Blue Prism diagram that orchestrates a business transaction by calling object actions and sub-processes; it holds no application-specific logic and is where queue items are controlled.

The reusable, application-facing tier containing business objects and their actions, which encapsulate spying, navigation, and data read/write for a single target application. Processes call these actions rather than interacting with applications directly.

A Blue Prism component that models one application or system and exposes actions the process can call. It bundles the application model, spied elements, and page logic needed to perform discrete operations on that system.

A single flowchart canvas inside a process or object. Pages hold stages and links, can be called from other pages, and are used to break logic into named, testable units such as initialise, main loop, or cleanup.

A visual building block placed on a page, such as Action, Decision, Calculation, Data Item, Loop, or Read. Each stage type performs a specific operation when the running token reaches it.

The directed arrow connecting stages that defines the execution path of the token. Links carry no data themselves and can be followed only when the source stage has completed.

The marker showing the currently executing position on a diagram. Execution advances token by token along links, and understanding token movement helps diagnose why a branch did or did not run.

A branching stage that evaluates a boolean expression and sends the token down the True or False outgoing link. All paths from a decision must eventually converge or end so the page cannot dead-end unexpectedly.

A multi-way branching stage that compares one expression against several named outcomes and routes the token down the matching link. It is often clearer than nested decisions when handling many discrete cases.

A stage that evaluates an expression and stores the result in a data item, or simply evaluates an expression for its side effect. It is the standard place for string building, arithmetic, and simple transformations.

A named variable declared on a page or at process/object level, with a defined data type and optional initial value. Its scope determines which pages and stages can read or write it.

A structured data item holding rows and fields, similar to a table. Collections store queue data, scraped results, and loop input, and are manipulated with collection manipulation actions such as filter, sort, and merge.

A stage that iterates over a collection or a fixed count, exposing the current row or index to the loop body. Each pass follows the loop link, and the loop ends when the collection is exhausted.

The wizard used to define an application model: launch and attach behaviour, target application type, and the element tree of windows and controls. Spied elements are stored here and referenced by read, write, and navigate stages.

The interactive technique for capturing interface elements from a running application into the object's element tree. Different spy modes suit different technologies, and choosing the wrong one produces brittle or unmatched elements.

A stored reference to a control or window in a target application, defined by attributes such as path, name, class, or text. Actions read from, write to, or click elements by name through these references.

The set of properties, such as window title, class, or control id, that Blue Prism uses to locate an element at runtime. Too many volatile attributes cause failures; too few cause the wrong element to be matched.

A pattern character, commonly an asterisk, used in element match attributes or text comparisons to stand for variable text. It makes matching tolerant of changing values such as document names or record identifiers.

A stage that performs an interface action on an element, such as click, focus, or key press, or that launches or attaches to an application. It is the standard way objects drive user interfaces.

A stage that extracts a value from an application element, or reads text from the screen, into a data item. Reads should be paired with waits or retries because interfaces may not update instantly.

A stage that pauses execution until a specified condition on an element is met, or until a timeout expires. It is preferred over fixed delays because it synchronises with the application rather than guessing timings.

Calling an external system through an HTTP endpoint instead of its user interface, typically by configuring a web service definition in Blue Prism and mapping request and response fields. It is generally faster and more stable than screen automation.

Driving applications delivered through a remote desktop or virtual channel where normal spying is unavailable, so interaction relies on surface automation, image matching, and keyboard navigation. Designs must tolerate slower, less reliable element recognition.

A technique for interacting with applications through screen coordinates, images, and keyboard input when the underlying interface cannot be spied. It is more fragile, so it needs careful waits, region definitions, and frequent validation.

The design of blocks, recover stages, and logic that respond when a stage fails or a business rule is violated. Good handling classifies the failure, records context, and decides whether to retry, skip, or stop.

A stage that marks where execution resumes after an exception is handled within a block. When an error occurs, the token jumps to the recover stage rather than the next linked stage.

A grouping of stages with a defined start and end that can be wrapped in exception handling. Blocks limit how far an error propagates and let the designer attach recovery logic to a specific section of work.

An error raised by the runtime, such as an element not found or an application crash, indicating the automation could not complete as designed. System exceptions typically trigger retry or item deferral logic.

An error deliberately raised when data or a situation violates a business rule, such as a missing mandatory field or an unrecognised account type. It is usually marked as a business exception on the queue item rather than retried indefinitely.

A Blue Prism container that holds items to be processed, with statuses, priorities, tags, and attempt counts. Queues decouple data intake from processing and let multiple resources share the workload safely.

A single record in a work queue, carrying key data, status, priority, and exception details. Its lifecycle moves through states such as new, locked, completed, and exception.

The mechanism that reserves a queue item for one resource while it is being processed, preventing two robots from working the same record. Locks are released when the item is completed, marked exception, or the session ends.

Returning a queue item to the queue with a future time so it can be retried later, often after a transient system problem. Deferrals keep the item available without blocking the resource that encountered the issue.

The Blue Prism management console for deploying processes, starting and stopping sessions, viewing queues, and monitoring logs. Administrators use it to observe runtime health and intervene when sessions fail.

A running instance of a process on a resource, which can be started manually or by a schedule. Each session has its own status, logs, and queue activity visible in the Control Room.

A configured rule that starts a process at defined times or intervals on specified resources, optionally with retry and end conditions. Schedules are the standard way to run unattended automation predictably.

A machine or runtime environment registered with Blue Prism that can host sessions. Resources are assigned to processes and pools, and their availability determines whether scheduled work can run.

The assignment of permissions through roles so users see and change only what their responsibilities require. Roles govern access to processes, queues, credentials, and administrative functions, supporting least privilege.

The secure store that holds usernames and passwords referenced by name in automation, so secrets never appear in process logic. Access to each credential is granted to specific roles and logged when used.

The record of configuration and security-relevant changes, such as permission edits, credential access, and process releases. It supports investigation, compliance evidence, and change accountability.

The controlled promotion of processes and objects between environments, using packaged releases and versioning so that tested content reaches production intact. It includes change approval and rollback planning.

A distinct Blue Prism database and runtime set, such as development, test, or production, used to separate stages of the delivery lifecycle. Promotion between environments should follow an approved, repeatable path.