$exec Directive | BizFirst Expression Evaluation

Property	Value
Directive name	`exec` (or `execFacts` in current implementation)
Required isolation level	Safe
Project	`FlowExecution.Services`
Data source	`INodeExecutionFacts`

Complete Path Reference

Path	Type	Example	Description
`$exec.executionId`	String (GUID)	`e8f23a1b-...`	Unique ID for this execution instance
`$exec.threadId`	String	`t_main`	Current execution thread identifier
`$exec.nodeKey`	String	`SendInvoiceEmail`	Key of the currently executing node
`$exec.startedAt`	String (ISO 8601)	`2024-03-15T10:22:01Z`	When this execution started
`$exec.triggeredBy`	String	`Webhook`	Trigger type: Manual / Webhook / Scheduled / Form
`$exec.triggeredByUserId`	Number (nullable)	`42`	ID of user who triggered (null for scheduled/webhook)
`$exec.retryCount`	Number	`0`	How many times this node has been retried
`$exec.isRetry`	Boolean	`false`	true if this is a retry attempt
`$exec.parentExecutionId`	String (nullable)	`d3a8...`	ID of the parent execution (set for sub-workflows)
`$exec.correlationId`	String (nullable)	`ORDER-2024-001`	Business correlation ID passed at trigger time

Complex Scenarios

Scenario 1 — Idempotency Key for External API Calls

Use the execution ID as an idempotency key to prevent duplicate submissions on retry:

HTTP header: Idempotency-Key

{@ $exec.executionId }-{@ $exec.nodeKey }

e8f23a1b-4c2d-...-SendPayment — unique per execution AND node

Scenario 2 — Skipping Expensive Operations on Retry

If a node is being retried, skip a slow data-fetch step that already succeeded:

// IfCondition: "Skip if retry"
{@ $exec.isRetry }
// true → skip to next node (data already fetched in first attempt)
// false → run the full data fetch

// More nuanced: only skip if retried more than once
{@ $js`return context.exec.retryCount > 1` }

Scenario 3 — Structured Log Entry with Correlation

// Structured audit log node field:
{
  "timestamp":     "{@ $ctx.now }",
  "executionId":   "{@ $exec.executionId }",
  "correlationId": "{@ $exec.correlationId @default:none }",
  "workflow":      "{@ $flow.current.name }",
  "node":          "{@ $exec.nodeKey }",
  "triggeredBy":   "{@ $exec.triggeredBy }",
  "userId":        {@ $exec.triggeredByUserId @default:null }
}

Scenario 4 — Sub-workflow Traceability

When a sub-workflow runs, it can reference its parent execution for full traceability:

// Sub-workflow creates a child execution record linking to parent:
POST /api/audit/execution
{
  "executionId":       "{@ $exec.executionId }",
  "parentExecutionId": "{@ $exec.parentExecutionId @default:root }",
  "workflowId":        "{@ $flow.current.id }",
  "tenantId":          {@ $ctx.tenant.id }
}

Scenario 5 — Time-Based Execution Reporting

Build a duration report by comparing execution start to current time:

{@ $js`
const started = new Date(context.exec.startedAt);
const now     = new Date(context.ctx.now);
const seconds = Math.round((now - started) / 1000);
return seconds < 60 ? seconds + 's' : Math.round(seconds/60) + 'm';
` }
// "47s" or "3m" — human-readable execution duration so far

Common Errors

Error	Cause	Fix
`PathNotFound: correlationId`	No correlation ID was passed at trigger time	Use `@default:none`; make correlation ID optional in your design
`PathNotFound: parentExecutionId`	This is a top-level execution, not a sub-workflow	Use `@default` to handle both top-level and sub-workflow contexts

$exec — Execution Directive