Architecture Patterns

Use this skill when the task is larger than one function or trigger. Pick the pattern from the requirement, then implement it with Function, Trigger, Worker, observers, state, queues, streams, and pub/sub.

Pattern Map

Requirement	Pattern	iii shape
Sequential work with retries, DLQ, step tracking	Durable workflow	Functions chained through named queues, progress in state
Keep views, metrics, cache, or clients in sync	Reactive backend	State triggers plus stream/pubsub side effects
React to another function's calls, results, or failures without changing it	Observer side effect	`iii.observe({ function_id, on, filters }, target)` (`iii-observe`)
Specialized AI agents hand work to each other	Agentic backend	One function per agent, queue handoffs, shared state
Commands publish events and projections update independently	Event-driven CQRS	Command functions, event log in state, subscribe triggers
Pure, traceable composition	Effect pipeline	Small functions composed synchronously with `trigger()`
Webhook/cron automation chains	Low-code automation	Trigger, transform, action nodes chained by enqueue

Durable Workflow

Use named queues when steps need different retry, FIFO, or concurrency policy.

TypeScript

import { registerWorker, TriggerAction } from "iii-sdk";

const iii = registerWorker("ws://localhost:49134", { workerName: "order-workflow" });

async function track(orderId: string, step: string, status: string) {
  await iii.trigger({
    function_id: "state::update",
    payload: { scope: "orders", key: orderId, ops: [{ type: "merge", path: ["steps"], value: { [step]: status } }] },
  });
}

iii.registerFunction("orders::validate", async (order) => {
  await track(order.id, "validate", "done");
  return iii.trigger({
    function_id: "orders::charge",
    payload: order,
    action: TriggerAction.Enqueue({ queue: "order-payment" }),
  });
});

iii.registerFunction("orders::charge", async (order) => {
  await track(order.id, "payment", "done");
  return iii.trigger({
    function_id: "orders::ship",
    payload: order,
    action: TriggerAction.Enqueue({ queue: "order-ship" }),
  });
});

Python

from iii import register_worker

iii = register_worker("ws://localhost:49134")

def track(order_id, step, status):
    iii.trigger({
        "function_id": "state::update",
        "payload": {
            "scope": "orders",
            "key": order_id,
            "ops": [{"type": "merge", "path": ["steps"], "value": {step: status}}],
        },
    })

def validate(order):
    track(order["id"], "validate", "done")
    return iii.trigger({
        "function_id": "orders::charge",
        "payload": order,
        "action": {"type": "enqueue", "queue": "order-payment"},
    })

iii.register_function("orders::validate", validate)

Rust

use iii_sdk::TriggerAction;
use iii_sdk::protocol::TriggerRequest;
use serde_json::json;

async fn enqueue_charge(iii: iii_sdk::IIIClient, order: serde_json::Value) -> Result<serde_json::Value, iii_sdk::Error> {
    iii.trigger(TriggerRequest {
        function_id: "state::update".into(),
        payload: json!({
            "scope": "orders",
            "key": order["id"],
            "ops": [{ "type": "merge", "path": ["steps"], "value": { "validate": "done" } }]
        }),
        action: None,
        timeout_ms: None,
    }).await?;

    iii.trigger(TriggerRequest {
        function_id: "orders::charge".into(),
        payload: order,
        action: Some(TriggerAction::Enqueue { queue: "order-payment".into() }),
        timeout_ms: None,
    }).await
}

Reactive Backend

Use state triggers when the requirement says "after create/update, do this", "avoid polling", or "push live updates".

iii.registerFunction("todos::on-change", async (event) => {
  await iii.trigger({
    function_id: "stream::send",
    payload: { stream_name: "todos-live", group_id: "default", data: event.new_value },
    action: TriggerAction.Void(),
  });
});

iii.registerTrigger({
  type: "state",
  function_id: "todos::on-change",
  config: { scope: "todos" },
});

Observer Side Effects

Use an observer when "after X runs (or fails), also do Y" and X should not know about Y: audit logs, alerts, projections, cache invalidation. Every function is observable since 0.25, and an observer failure never affects the observed call.

import type { Observed } from "iii-sdk";

iii.registerFunction("audit::order-created", async (event: Observed) => {
  await iii.trigger({
    function_id: "state::set",
    payload: { scope: "audit", key: event.invocation_id, value: { order: event.input, by: event.caller?.worker_name } },
  });
  return null;
});

// every successful orders::create, except the replay tool's own calls
iii.observe(
  { function_id: "orders::create", filters: { "caller.worker_name": { $ne: "order-replayer" } } },
  "audit::order-created",
  { max_in_flight: 8 },
);

def notify(event):  # the Observed envelope: input, error, caller, ...
    print("charge failed", event["input"], event["error"]["code"])

iii.register_function("alerts::payment-failed", notify)
iii.observe(
    {"function_id": "payments::charge", "on": "error", "filters": {"input.amount": {"$gte": 500}}},
    "alerts::payment-failed",
)

Activations are at most once and fire-and-forget. When the side effect must complete with retries, have the target enqueue the work (TriggerAction.Enqueue). Filter on caller or caller.parent to stop echo loops; the engine also drops activations deeper than observers.max_observe_depth (8 by default).

Agentic Backend

Model each agent as a function with one responsibility. Store shared context in state and hand off work through named queues.

iii.registerFunction("agents::researcher", async (task) => {
  await iii.trigger({
    function_id: "state::set",
    payload: { scope: "research", key: task.id, value: { findings: [] } },
  });
  return iii.trigger({
    function_id: "agents::critic",
    payload: task,
    action: TriggerAction.Enqueue({ queue: "agent-tasks" }),
  });
});

Event-Driven CQRS

Commands validate and publish domain events. Projections subscribe independently and write query-optimized state.

iii.registerFunction("cmd::add-inventory-item", async (input) => {
  const event = { type: "inventory.item-added", itemId: input.itemId, quantity: input.quantity };
  await iii.trigger({
    function_id: "state::set",
    payload: { scope: "inventory-events", key: `${Date.now()}-${input.itemId}`, value: event },
  });
  await iii.trigger({ function_id: "publish", payload: { topic: event.type, data: event } });
  return { accepted: true };
});

iii.registerTrigger({
  type: "subscribe",
  function_id: "proj::inventory-list",
  config: { topic: "inventory.item-added" },
});

// Alternative without pub/sub: the projection observes the command function directly and
// receives its input (the command) on every successful call.
iii.observe({ function_id: "cmd::add-inventory-item", deliver: "input" }, "proj::inventory-list");

registerTrigger is deprecated in 0.25 (one warning per process) but stays the way to bind the built-in subscribe, state, cron, http, and queue providers, which have not migrated to source functions yet.

Selection Rules

Use sync composition for short effect pipelines where the caller needs the final result.
Use enqueue for unreliable, slow, or must-complete steps.
Use pub/sub for independent fan-out where each subscriber can tolerate event-style delivery.
Use state triggers for derived views and side effects that should run automatically after writes.
Use observers for side effects of a specific function's calls, results, or failures, when the caller should not know about them and at-most-once delivery is acceptable.
Use cron triggers for scheduled maintenance and periodic starts.
Keep each function small enough to test independently.

When to Use

Use this skill for complete backend patterns: workflows, agentic systems, reactive apps, CQRS, effect pipelines, and automation chains.
Use this when the request describes product behavior rather than a single SDK API call.

Boundaries

For exact trigger config, function registration syntax, custom triggers, channels, and HTTP-invoked functions, use iii-core-primitives.
For observer options, filters, and owner-delivered event sources, use iii-observe.
For engine ports, adapters, engine-owned workers, and RBAC (rbac-proxy), use iii-engine-config; queue retry and FIFO policy live with the queue worker's docs.
For package-specific SDK syntax, use iii-sdk-reference.
Worker-backed capability details live with the worker docs, not as top-level iii skills.

iii-architecture-patterns