Custom Triggers

Comparable to: Custom event adapters, webhook receivers

Key Concepts

Use the concepts below when they fit the task. Not every custom trigger needs all of them.

• registerTriggerType({ id, description }, handler) defines a new trigger type with registerTrigger and unregisterTrigger callbacks
• The handler receives a TriggerConfig containing id, function_id, config, and optional metadata
• When the external event fires, call iii.trigger({ function_id, payload: event }) to invoke the registered function
• unregisterTriggerType cleans up when the trigger type is no longer needed
• Do not reuse built-in trigger type names: http, cron, durable:subscriber, state, stream, subscribe

Architecture

External event source (webhook, file watcher, IoT, CDC, etc.)
  → Custom trigger handler (registerTriggerType)
    → iii.trigger({ function_id, payload: event })
      → Registered function processes the event

iii Primitives Used

Reference Implementation

See ../references/custom-triggers.js for the full working example — a custom trigger type that listens for external events and routes them to registered functions.

Also available in Python: ../references/custom-triggers.py

Also available in Rust: ../references/custom-triggers.rs

Common Patterns

Code using this pattern commonly includes, when relevant:

• registerWorker(url, { workerName }) — worker initialization
• Node: iii.registerTriggerType({ id, description }, { registerTrigger, unregisterTrigger })
• Python: iii.register_trigger_type({ "id": id, "description": description }, TriggerHandlerSubclass())
• Rust: iii.register_trigger_type(RegisterTriggerType::new(id, description, handler))
• registerTrigger(config) / register_trigger(config) is invoked by the engine when a trigger instance is registered
• unregisterTrigger(config) / unregister_trigger(config) is invoked when that trigger instance is removed
• iii.trigger({ function_id: config.function_id, payload: eventPayload }) — fire the target function
• Cleanup logic in unregisterTrigger (close connections, remove listeners, clear intervals)
• const logger = new Logger() — structured logging

Authoring Guidance

• Store trigger bindings in a map keyed by config.id.
• Keep long-lived resources per trigger binding: sockets, timers, webhook routes, subscriptions, or file watchers.
• Put subscription/listener setup in registerTrigger, not in module top-level code.
• Put cleanup in unregisterTrigger so deleted triggers do not leak timers or connections.
• Put static event-source settings in trigger config.
• Put event-specific data in the call payload sent to iii.trigger().
• If using condition_function_id, let the trigger dispatch path evaluate the condition before invoking the target function.
• Add trigger_request_format and call_request_format schemas where the SDK supports them, so discovery and generated skills know the config and handler payload shapes.

Adapting This Pattern

Use the adaptations below when they apply to the task.

• Choose a unique trigger type name that describes your event source (e.g. file-watcher, mqtt, db-cdc)
• In registerTrigger, start the listener (open socket, receive webhook, subscribe to topic, attach watcher, or consume CDC)
• Use polling only as a fallback when the external source has no webhook, subscription, stream, or CDC mechanism.
• In unregisterTrigger, tear down the listener to avoid resource leaks
• Store active listeners in a map keyed by config.id for clean unregistration
• Pass relevant event data in the payload when calling iii.trigger({ function_id, payload: event })

Pattern Boundaries

• If the task uses built-in HTTP routes, prefer iii-http-endpoints.
• If the task uses built-in cron schedules, prefer iii-cron-scheduling.
• If the task uses built-in queue triggers, prefer iii-queue-processing.
• Stay with iii-custom-triggers when iii has no built-in trigger type for the event source.

When to Use

• Use this skill when the task is primarily about iii-custom-triggers in the iii engine.
• Triggers when the request directly asks for this pattern or an equivalent implementation.

Boundaries

• Never use this skill as a generic fallback for unrelated tasks.
• You must not apply this skill when a more specific iii skill is a better fit.
• Always verify environment and safety constraints before applying examples from this skill.