System Triggers

System triggers are the foundation of ambient agents. They initiate workflows based on events, changes, or schedules so work moves forward automatically.

Trigger types

TriggerSummaryResponsivenessTypical objectiveExample
WebhooksEvent-driven; fire in real time when a specific event occurs in a source systemInstantaneousRespond to key business events“Gong call completed” → extract key details and propose CRM updates
PollersPull-based; periodically check systems that don’t support webhooksNear-instantaneousDetect important changesAccount health drops from Green → generate get-well plan
ScheduledRun at predefined times or intervalsPeriodicKeep time-based workflows movingEvery Monday 9am → review feature requests and route to PM

📘

Availability

Webhooks are in Limited Preview; Scheduled and Pollers are coming soon.

How triggers connect to your processes

In Agent Studio, a Plugin consists of three components: a Trigger (when to run), a Launch Configuration (who can run it), and a Process (what it does). System triggers define the "when to run" condition, delivering events to your Process so the agent can reason and act.

Webhooks at a glance (Limited Preview)

  • Create a Listener to capture events from providers (e.g., Salesforce, DocuSign, Asana).
  • Moveworks validates incoming events, supporting signed secrets and one-time challenge verification.
  • Events are parsed and dispatched to matching plugins, with observability provided through dedicated logs.

For hands-on steps, see Webhook Triggers (Limited Preview) and the Quickstart Guide.

Choosing the right trigger

  • Select Webhooks for critical, real-time workflows.
  • Choose Pollers for near-real-time reactions when source systems lack webhook support.
  • Opt for Scheduled triggers for cadence-based tasks and periodic audits.

Execution flow (high level)

  1. Trigger Activates: An event, change, or schedule initiates the process.
  2. Moveworks Processes the Trigger: Validates and parses the trigger (e.g., a webhook).
  3. Matching Plugin(s) Execute: The configured Process runs, using LLM reasoning, API calls, and optional approvals.
  4. Logs Provide Traceability: Monitor each stage end-to-end during development and operations.