AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents
TypeScript ai-agentai-agent-toolsai-agentsai-agents-frameworkmcpmcp-servermcp-toolsmcpsn8n-alternativeno-code-automationworkflowworkflow-automationworkflows
import { isNil } from '@activepieces/core-utils'
import { RunTimeline } from '@activepieces/shared'
// Assembles the leg-0 latency breakdown from the run timestamps + worker-measured phase durations.
// Pure and dependency-free so the phase math stays unit-testable in isolation.
// QUEUE = the remainder of (startTime - created) once provision and boot are removed.
// RUN = finishTime - startTime (matches the run's displayed "Took"); falls back to the worker's
// sandboxRun wall clock (runMs) until finishTime lands.
// Returns undefined (skip the write) until all inputs are present, and never overwrites an existing
// timeline so a resumed run keeps its first leg.
export function buildRunTimeline({
existingTimeline,
created,
startTime,
finishTime,
provisionMs,
bootMs,
runMs,
}: BuildRunTimelineParams): RunTimeline | undefined {
if (!isNil(existingTimeline)) {
return undefined
}
if (isNil(provisionMs) || isNil(bootMs) || isNil(runMs)) {
return undefined
}
if (isNil(created) || isNil(startTime)) {
return undefined
}
const startupMs = new Date(startTime).getTime() - new Date(created).getTime()
const queueMs = Math.max(0, startupMs - provisionMs - bootMs)
const runDurationMs = isNil(finishTime)
? runMs
: Math.max(0, new Date(finishTime).getTime() - new Date(startTime).getTime())
return {
legs: [[
{ name: 'QUEUE', durationMs: queueMs },
{ name: 'PROVISION', durationMs: provisionMs },
{ name: 'BOOT', durationMs: bootMs },
{ name: 'RUN', durationMs: runDurationMs },
]],
}
}
type BuildRunTimelineParams = {
existingTimeline?: RunTimeline | null
created?: string | null
startTime?: string | null
finishTime?: string | null
provisionMs?: number
bootMs?: number
runMs?: number
}