AWS Amazon Simple Workflow Service

Name: Amazon Simple Workflow Service API
Brand: Amazon Simple Workflow Service
Availability: InStock

★ Only Publicly Available OpenAPI DocumentDeveloper ToolsCi CdapiKey37 EndpointsREST

For Agents

Coordinate long-running, multi-step workflows by registering domains and workflow types, polling for activity and decision tasks, and signalling running executions.

Install Jentic One Beta

Connect the Amazon Simple Workflow Service to your agent

Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Amazon Simple Workflow Service, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.

Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.

Step 1: Jentic One Host machine

# On the machine that will host your Jentic One instance:
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh

Step 2: Agent machine

# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh
jentic register       # connects your agent to your Jentic One instance

Jentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.

Capabilities

What an agent can do with Amazon Simple Workflow Service API.

Start, signal, terminate, and request cancellation of workflow executions

Poll for decision tasks via PollForDecisionTask and respond with decisions

GET STARTED

Start building with Amazon Simple Workflow Service API

Explore with Jentic

View OpenAPI Document

Jentic publishes the only available OpenAPI specification for Amazon Simple Workflow Service, keeping it validated and agent-ready. Amazon Simple Workflow Service (SWF) coordinates work across distributed application components. It models long-running business processes as workflow executions made up of activity tasks dispatched to workers and decision tasks dispatched to a decider, with full execution history and at-most-once task delivery. The 37 operations cover domain and type registration, polling for tasks, recording activity heartbeats, signalling and cancelling executions, and counting open or closed workflows for capacity planning.

Use Cases

Patterns agents use Amazon Simple Workflow Service API for, with concrete tasks.

★ Long-Running Order Fulfilment Workflows

Model an e-commerce order as an SWF workflow execution that spans payment authorisation, inventory reservation, shipping label generation, and customer notification — each implemented as an activity task picked up by a worker. SWF guarantees at-most-once task delivery and retains execution history for replay, making it well suited to processes that run for hours or days and need a precise audit trail.

Call StartWorkflowExecution in domain 'orders', workflowType OrderFulfilment v1, workflowId order-2026-06-09-001, taskList orders-decider, with input containing the order JSON.

Human-in-the-Loop Approval Pipelines

Run approval workflows that pause until a human signals a decision. The decider issues a ScheduleActivityTask for the human reviewer's queue; when the reviewer responds, an external system calls SignalWorkflowExecution to deliver the decision. SWF holds execution state for up to a year, so workflows can wait days for review without consuming compute.

Call SignalWorkflowExecution with domain 'approvals', workflowId loan-app-2026-06-09-77, signalName ReviewerDecision, and input '{"decision":"approved","reviewer":"jane.doe"}'.

Workflow Capacity and Health Monitoring

Operate SWF at scale by polling CountOpenWorkflowExecutions and CountClosedWorkflowExecutions per domain and workflow type. Combine with ListOpenWorkflowExecutions to find stuck runs and GetWorkflowExecutionHistory to drill into the event log when an execution stalls. This is the standard observability pattern for SWF before resorting to custom CloudWatch metrics.

Call CountOpenWorkflowExecutions for domain 'orders' with executionFilter on workflowId prefix 'order-', and return the open count.

Agent-Orchestrated Workflow Control via Jentic

Let an AI agent kick off, signal, and cancel SWF executions through Jentic without holding AWS keys. The agent submits an intent like 'start a fulfilment workflow for order 1234'; Jentic loads StartWorkflowExecution, validates the input, and executes it. Subsequent agent actions (signal, cancel) flow through the same vaulted credentials, keeping the agent context free of secrets.

Through Jentic, search for 'start an aws swf workflow execution', load StartWorkflowExecution, and execute with domain, workflowId, workflowType, and input.

Key Endpoints

37 endpoints — jentic publishes the only available openapi specification for amazon simple workflow service, keeping it validated and agent-ready.

METHOD

PATH

DESCRIPTION

POST

/#X-Amz-Target=SimpleWorkflowService.StartWorkflowExecution

Start a workflow execution

POST

/#X-Amz-Target=SimpleWorkflowService.PollForDecisionTask

Poll for the next decision task

POST

/#X-Amz-Target=SimpleWorkflowService.PollForActivityTask

Poll for the next activity task

POST

/#X-Amz-Target=SimpleWorkflowService.SignalWorkflowExecution

Signal a running workflow

POST

/#X-Amz-Target=SimpleWorkflowService.GetWorkflowExecutionHistory

Read full execution history

POST

/#X-Amz-Target=SimpleWorkflowService.RecordActivityTaskHeartbeat

Heartbeat a long-running activity

POST

/#X-Amz-Target=SimpleWorkflowService.StartWorkflowExecution

Start a workflow execution

POST

/#X-Amz-Target=SimpleWorkflowService.PollForDecisionTask

Poll for the next decision task

POST

/#X-Amz-Target=SimpleWorkflowService.PollForActivityTask

Poll for the next activity task

POST

/#X-Amz-Target=SimpleWorkflowService.SignalWorkflowExecution

Signal a running workflow

POST

/#X-Amz-Target=SimpleWorkflowService.GetWorkflowExecutionHistory

Read full execution history

Why Jentic?

Three things that make agents converge on Jentic-routed access.

Credential isolation

AWS access keys for Amazon Simple Workflow Service (used to compute the AWS Signature v4 HMAC in the Authorization header) are held encrypted in the Jentic vault. Agents receive scoped, short-lived execution permissions; the raw access key ID and secret access key never enter the agent's prompt or memory.

Intent-based discovery

Agents call Jentic with an intent like 'start an aws swf workflow execution' and Jentic returns the matching Amazon Simple Workflow Service operation along with its input schema, so the agent can invoke the right action without parsing AWS service docs.

Time to first call

Direct integration with Amazon Simple Workflow Service: 2-4 days to wire up SigV4 signing, error handling, retries, and IAM scoping. Through Jentic: under an hour — search by intent, load the operation schema, execute.

FAQs

Specific to using Amazon Simple Workflow Service API through Jentic.

Why is there no official OpenAPI spec for Amazon Simple Workflow Service?

AWS does not publish an OpenAPI specification for Amazon Simple Workflow Service; it ships Smithy models and language-specific SDKs instead. Jentic generates and maintains this OpenAPI spec so that AI agents and developers can call Amazon Simple Workflow Service via structured tooling. It is validated against the live API and kept up to date. Get started at https://app.jentic.com/sign-up.

What authentication does Amazon Simple Workflow Service use?

SWF uses AWS SigV4 in the Authorization header with X-Amz-Target identifying the action (e.g. SimpleWorkflowService.StartWorkflowExecution). Through Jentic, your AWS access keys are stored encrypted in the vault and Jentic computes the SigV4 signature for each request.

Can I orchestrate human-in-the-loop steps with SWF?

Yes. Pause an execution by scheduling a long-timeout activity for a human-task queue, or have the decider stop scheduling new tasks until an external SignalWorkflowExecution call delivers the human decision. SWF retains execution state for up to a year.

What are the rate limits for Amazon SWF?

Default per-account limits include 100 starts per second per workflow type, 200 polls per second per task list, and a 1MB execution input/output size cap. Heartbeats and decisions have separate per-task-list quotas. Most limits are increasable via AWS Support.

How do I start an SWF workflow execution through Jentic?

Search Jentic with 'start an aws swf workflow execution', load the StartWorkflowExecution schema, and execute with domain, workflowId, workflowType (name + version), taskList, and an input string. Jentic returns the runId of the new execution.

Should I use SWF or AWS Step Functions for new workflows?

For new projects AWS recommends Step Functions, which offers a managed state machine and visual editor. Use SWF when you have an existing decider/activity-worker codebase, need finer control over task dispatch, or rely on its long execution retention.

Can I cancel a running workflow execution?

Yes. RequestCancelWorkflowExecution sends a cancellation request that the decider can act on; TerminateWorkflowExecution forces an immediate termination without giving the decider a chance to clean up.

AWS Amazon Simple Workflow Service

Connect the Amazon Simple Workflow Service to your agent

Capabilities

Start building with Amazon Simple Workflow Service API

Use Cases

Key Endpoints

Why Jentic?

Related APIs

FAQs