# Using Amazon Quick Flows
Amazon Quick Flows is a capability within Amazon Quick that lets any user create, customize, and share workflows that automate routine tasks. You can generate flows from conversations with chat agents, describe what you need in natural language, or build them manually using the visual editor — no technical skills required. Flows can also be published to an admin-managed library and shared with other Amazon Quick users in your organization.
Each flow is a sequence of steps that can gather user input, generate AI responses from your data or the web, take actions in connected applications, and apply logic to control how steps run.
## What are Flows?
A flow is a sequence of *steps* that you define once and run whenever you need it. You build flows in the editor, then run them in either a guided step-by-step mode or a conversational chat mode where you can refine outputs and ask follow-up questions.
Each flow can include the following step types:
### AI responses
+ Chat agent step that gets a response from a custom agent and can take actions.
+ Research step that conducts research on a topic.
+ Web search step that gets a response using web search.
+ General knowledge step that gets a response directly from base AI models.
+ UI agent step to perform tasks on public websites.
+ Create image step that generates an image from inputs.
### Flow logic
+ Reasoning group step that adds run instructions to one or more steps.
### Data insights
+ Quick data step that gets insights from spaces or knowledge bases.
+ Dashboards and topics step that gets responses from dashboards and topics.
### Actions
+ Application actions step that reads or writes to connected apps.
### User input
+ Text step to get and use text from users.
+ Files step to get and use files from users.
For detailed explanations of each component, see [Flow components and features](flow-components-and-features.md).
## Why Flows?
Organizations face business processes that require both human judgment and system interactions. Amazon Quick Flows bridges this gap by combining AI reasoning with direct business actions, so you can automate tasks like generating responses for requests for proposals (RFPs), reviewing statements of work (SOWs), or compiling industry trends into a sales pitch.
Amazon Quick Flows can help across departments:
+ *Sales, marketing, and operations*: Qualifying leads, generating personalized proposals, creating marketing content, updating customer relationship management (CRM) records, and supporting processes like RFP responses
+ *Human resources (HR)*: Processing employee requests, answering policy questions, and automating onboarding steps
+ *Finance*: Analyzing expense reports, flagging anomalies, and processing routine approvals
+ *Information technology (IT)*: Automating troubleshooting, system monitoring, and access management
# Getting started with Quick Flows
This section covers user personas, prerequisites, and the Amazon Quick Flows landing page interface.
# User personas
Amazon Quick Flows serves three types of users:
## Creator
Creators are business users who build flows. They understand the repetitive tasks and processes in their domain and translate them into reusable workflows.
## End user
End users run flows that creators have built. They interact with flows to complete their daily tasks and provide feedback on how well the automation works.
## Admin
Admins configure and manage governance controls for Amazon Quick Flows within Amazon Quick. They can:
+ Enable or disable Quick Flows for the account
+ Restrict access to Quick Flows for specific users
+ Enable or disable Bedrock models for output refinement
+ Require approval before users can share flows
+ Unpublish shared flows and transfer ownership between users
# Prerequisites for Quick Flows
Before you can create and use Amazon Quick Flows, you need to ensure that your Amazon Quick administrator has completed the following prerequisites.
## Administrator setup requirements
Your Quick administrator must complete the following tasks before you can create and use Quick Flows:
+ Set up and configure Quick for your organization. For more information, see [Setting up and signing into Amazon Quick](setting-up.md).
+ (Optional) Restrict access to flows for specific users using custom permissions.
For browser and region requirements, see [Supported browsers](supported-browsers.md).
## Required permissions
Permissions to create, run, share, and govern flows are determined by user subscriptions and any configured custom permissions. For more information, see [Managing Quick subscriptions](managing-subscriptions.md) and [Managing user access inside Amazon Quick](managing-users.md).
## Amazon Bedrock model access
Quick Flows uses Amazon Bedrock models for AI reasoning in the General knowledge step. Your administrator must enable access in custom permissions for output refinement in flows using Bedrock models. For more information, see [General knowledge](ai-response-steps.md#general-knowledge-step).
## Next steps
After ensuring that all prerequisites are met, you can:
+ Learn about key concepts in Quick Flows. See [Terminology and key concepts](terminology-and-key-concepts.md).
+ Create your first flow. See [Creating flows](creating-flows.md).
# Amazon Quick Flows landing page
When you access Amazon Quick Flows, you are directed to the landing page. The landing page has three navigation tabs:
+ **Flows library** — Browse, search, and manage your flows and flows that others have shared with you.
+ **My schedules** — Monitor and manage scheduled flow runs.
+ **Pending approval** — Review and act on flows that are awaiting approval.
## Flows library
The Flows library organizes your flows into three views:
+ **Recently used** — Flows you have recently opened or modified, sorted by most recent.
+ **My Flows** — Flows you own or that have been shared with you directly.
+ **All Flows** — Every flow available in your organization.
Use the search bar to find flows by name or description. For each flow, the context menu lets you open, duplicate, or delete it. Choose **Create flow** to start building a new flow. For more information, see [Creating flows](creating-flows.md).
# Using pre-built flows
Amazon Quick Flows provides pre-built flows and samples to help you get started.
## Pre-built flows
Pre-built flows are ready-to-use templates available in **My Flows** on the Flows landing page. You can run them as-is or duplicate and customize them to fit your needs.
+ **Blog post generator** — Creates blog content based on a topic, target audience, and key points you provide. The flow produces structured posts with headings, body content, and suggestions for tone and style.
+ **Email composer** — Drafts professional emails for common business scenarios such as announcements, requests, and follow-ups. You specify the recipient context and purpose, and the flow generates appropriately toned content with a subject line.
+ **Data summarizer** — Analyzes structured data such as CSV or tabular files and extracts key insights, trends, and anomalies. The flow presents findings in a summary format you can share or use in other steps.
+ **Chart creator** — Generates visualizations based on your data and analysis goals. The flow recommends chart types based on your data and lets you configure the output for clarity.
+ **Meeting summarizer** — Processes meeting notes or transcripts and extracts discussion topics, decisions, and action items. The flow organizes the output into a structured summary you can distribute to attendees.
+ **Project status reporter** — Creates project status updates from key metrics and milestones you provide. The flow highlights risks, resource allocation, and next steps in a formatted report.
## Flow samples
Flow samples are available on the **Create Flow** page. These are example prompts designed to help you construct your own flows. Each sample provides a detailed natural language description that you can use as a starting point.
For example, the "Find current industry trends and news" sample generates a prompt like:
"Help me prepare for my customer call by fetching latest news on the company. Accept customer name, industry, company size, meeting objective, and previous interaction notes as inputs. For the customer name and previous interaction notes, search from web to obtain key talking points and relevant industry trends. Finally, from the web analysis and previous interaction notes, generate a meeting brief that I can use as talking points along with relevant product recommendations from indexed data."
## Customizing a pre-built flow
To customize a pre-built flow, duplicate it and modify the copy:
1. Select the pre-built flow you want to customize.
1. Choose **Duplicate** to create your own copy.
1. Open the duplicated flow in the Flow editor.
1. Modify the steps, prompts, and configuration as needed.
1. Save your customized flow.
For more information about editing flows, see [Editing flows](editing-flows.md).
# Flow components and features
These topics provide detailed information about individual step types and their capabilities. For configuration instructions, see [Editing flows](editing-flows.md).
# Terminology and key concepts
Understanding the core terminology and concepts helps you effectively create, run, share, and maintain flows within your organization.
## Steps and @ references
A flow is made up of steps that each perform a specific function, such as calling an action, querying your data, or searching the web. Most steps are controlled through a natural language prompt.
Steps pass data to each other using @ references. When you write a prompt inside a step, type @ to see a menu of previous steps. Select one to include that step's output as context in your prompt. For example, if Step 1 collects a customer's issue and Step 3 needs to classify it, Step 3's prompt might say: "Classify the issue in @Customer Issue by severity (low, medium, high)."
The available step types are organized into the following groups:
AI responses
+ **Chat agent** — Gets a response from a custom agent and can take actions in connected applications.
+ **Research** — Invokes Amazon Quick Research to generate research reports as part of your workflow.
+ **Web Search** — Generates responses using internet search results.
+ **General knowledge** — Gets a response directly from Amazon Bedrock models, with configurable response preferences for speed or versatility.
+ **UI Agent** — Navigates public websites that do not require a login and performs tasks like scrolling to find information or filling forms.
+ **Create Image** — Generates AI images from text inputs.
Flow logic
+ **Reasoning Group** — Groups related steps together with natural language instructions that define conditions, loops, validation, and execution order.
Data insights
+ **Quick data** — Retrieves responses from spaces and knowledge bases.
+ **Dashboards and topics** — Gets insights from Amazon Quick Sight dashboards and topics.
Actions
+ **Application actions** — Performs read or write operations in connected third-party applications through pre-built connectors.
User input
+ **Text** — Collects free-form text input from users.
+ **Files** — Accepts file uploads from users for document processing.
## Editor and Run mode
Editor mode is where you build your flow. You see all the steps laid out and can select each one to change its configuration. Run mode is where you test and execute your flow, with a chat panel where you can ask follow-up questions or refine the output.
# AI response steps
AI response steps generate content using AI models. Amazon Quick Flows provides the following AI response step types.
## Chat agent
Amazon Quick Flows allows you to use your chat agents to generate outputs from configured spaces or take action with configured action integrations, all within a workflow step.
Chat agents contain domain-specific knowledge, custom instructions, and connected tools. When you integrate a chat agent into a flow, you can automatically apply this specialized knowledge across multiple workflows without recreating it. For example, if you built a sales assistant chat agent that understands product details and follows brand guidelines, you can embed it in your outreach flow to ensure consistent communication at scale.
For configuration instructions, see [Editing flows](editing-flows.md).
**Note**
The chat agent step is a single-turn interaction. The agent responds to the task you instruct it to do, but does not support a back-and-forth conversation within the same step.
## Research
The research step invokes Amazon Quick Research to generate research reports within your flow. This lets you embed research directly into multi-step workflows — for example, creating account plans, conducting policy reviews, researching patent prior art, or generating industry reports.
For full details about Quick Research capabilities and limitations, see [Using Amazon Quick Research](using-amazon-quick-research.md). For configuration instructions, see [Editing flows](editing-flows.md).
You can reference the research output in later steps — for example, to send a summary over email to your team.
## Web search
The web search step lets your flows retrieve current information from the internet. This is useful when you need to access real-time data, verify facts, or gather information from public sources beyond your organization's internal knowledge base.
Write a prompt describing what to search for. The search results can be referenced by later steps in your flow using @ references.
For configuration instructions, see [Editing flows](editing-flows.md).
**Note**
Search results may vary over time as internet content changes. Some content may not be accessible through web search.
## General knowledge
The General knowledge step generates text responses using Amazon Bedrock models. Instead of selecting a specific model, you choose a response preference, and Amazon Quick Flows automatically selects the most appropriate model based on your preference and the requirements of your flow.
Choose from:
+ **Fast responses** — Optimized for speed across image, video, and text inputs.
+ **Versatility and performance** — Balanced capabilities for diverse tasks.
Optionally adjust the creativity slider to control the randomness of the response.
If you do not see response preferences, verify that your administrator has enabled "Enable bedrock model usage in General knowledge step for output refinement" in the Custom Permissions page.
For configuration instructions, see [Editing flows](editing-flows.md).
## UI agent
The UI agent step (Preview) lets your flows interact with public websites that do not require a login. The agent can autonomously navigate websites, click, type, read data, and produce structured outputs — all described in natural language.
**Writing effective instructions**
+ Be clear and specific about the task you want performed.
+ Use single, complete URLs (for example, "Go to https://example.com/reports").
+ Add constraints to narrow the scope (for example, "only look at the pricing section").
+ Specify when the agent should stop (for example, "stop after finding the first matching result").
+ Define the output format if needed (for example, "return the data as a bulleted list").
For configuration instructions, see [Editing flows](editing-flows.md).
**Note**
UI agent is currently in Preview. Some websites implement anti-automation measures such as CAPTCHA challenges that may limit UI agent capabilities. Websites that require login are not currently supported.
## Create Image
The Create Image step generates AI images from text prompts. You can configure creativity level, exclude terms, and image seed in the advanced settings.
For configuration instructions, see [Editing flows](editing-flows.md).
# Flow logic steps
Flow logic steps control how your flow runs.
## Reasoning Group
Reasoning groups give you control over how parts of your flow run using natural language instructions. A reasoning group contains its own set of steps — like an isolated workflow within your larger workflow — that runs based on conditions you define. You can add most step types to a reasoning group, except reasoning groups and research steps. Templates are available to help you get started.
### Loops
You can repeat the steps in a group for each value in a list from a previous step's output. Reference the previous step in your instructions, and the Flows runtime handles the iteration for you. For example, if a previous step returns a list of customer emails, a reasoning group can process each email in turn.
### Conditions
You can run the steps in a group based on natural language conditions that evaluate a previous step's output. For example, "Run if @Customer Priority is HIGH PRIORITY" routes only urgent items through the group's steps.
### Validation
You can check inputs or outputs before proceeding. For example, a reasoning group can verify that a required field is present before passing data to an action step.
For configuration instructions, see [Editing flows](editing-flows.md). For reasoning group limits, see [Quick Flows limits](quick-flows-limits.md).
# Data insight steps
Data insight steps retrieve information from your Quick data sources.
## Quick data
The Quick data step retrieves responses from your spaces and knowledge bases. Write a prompt describing what content to retrieve, and optionally link specific resources. By default, responses are generated from all knowledge sources the user has access to.
The system searches across indexed documents in your spaces, including knowledge bases from connected sources. For more information about setting up spaces and knowledge bases, see [Working with integrations](working-with-integrations.md).
For configuration instructions, see [Editing flows](editing-flows.md).
## Dashboards and topics
The Dashboards and topics step generates insights from your existing Amazon Quick Sight dashboards and topics. Responses can include charts, graphs, tables, and other visualizations.
Select a Quick Sight source (Dashboard or Topic) and write a prompt describing the insights you want. You can specify filters, date ranges, and other criteria in natural language.
For configuration instructions, see [Editing flows](editing-flows.md).
# Action steps
Action steps let your flows perform read or write operations in connected applications. The available operations depend on the action integrations that you have configured or that have been shared with you. For prerequisites, authentication methods, and available integrations, see [Working with integrations](working-with-integrations.md).
## Action parameters
Some actions require or accept parameters such as filters or specific input values. To determine which parameters an action accepts, you can ask Quick — for example, "What parameters does list emails accept when filtering emails?"
Parameters can be:
+ Supplied directly in the prompt (for example, "List emails with subject containing 'quarterly report'")
+ Referenced from a previous step using @ references (for example, a user input step that collects a search term)
+ Acquired dynamically during flow execution based on the flow context
## Working with list results
When you use an action that returns a list of items (for example, listing emails or tickets), the results may not include every item or every detail. This is because many applications return results in batches rather than all at once, and Quick retrieves only the first batch to keep your flow fast and responsive.
If you need the full details for each item in a list, you can use a reasoning group to go through the results one at a time and retrieve the complete information for each. For example, you might list your open support tickets in one step, then use a reasoning group to get the full details of each ticket.
When doing this, be mindful of how many items you are processing. A large number of items means more steps to run, which increases the time your flow takes to complete and the amount of data in your results. Where possible, use filters in your initial list action to narrow down the results before processing them.
For configuration instructions, see [Editing flows](editing-flows.md).
# User input steps
User input steps collect information from users when they run a flow.
## Text
The text input step collects free-form text from users. You can set a placeholder, a default value, and allow users to override the default at runtime.
Use placeholder text to guide users on what to enter. For example, you can present a set of options like "Enter 1 for Sales, 2 for Marketing, 3 for Support" to help users provide structured input that your flow can act on.
For configuration instructions, see [Editing flows](editing-flows.md).
## Files
The file upload step accepts a document, image, or video from users. You can upload a default file and allow users to override it at runtime. You can upload one file per step.
File uploads are subject to the same size and format restrictions as uploading files in chat. If your content exceeds these limits, consider using a space or knowledge base to process the request instead.
For configuration instructions, see [Editing flows](editing-flows.md).
# Creating and managing flows
This section covers how to create, edit, version, and share flows within your organization.
# Creating flows
You can create flows using natural language, the visual editor, by duplicating an existing flow, or from a chat agent conversation.
## Prerequisites for creating flows
Before creating a flow, ensure that you meet the requirements described in [Prerequisites for Quick Flows](prerequisites-for-quick-flows.md).
## Creating a flow using natural language prompt (NLP)
The natural language prompt method lets you describe your desired flow in plain language, and Amazon Quick Flows generates a flow based on your description.
1. Sign in to the Quick console.
1. In the navigation pane, choose **Flows**.
1. Choose **Create Flow**. You see a page where you can enter a natural language prompt or use a sample prompt.
1. In the prompt field, describe the flow you want to create. Be as specific as possible about:
+ The purpose of the flow
+ The inputs it should accept
+ The processing it should perform
+ The outputs it should produce
+ The actions it should take
1. Choose **Generate Flow**.
1. Review the generated flow and make any necessary adjustments.
1. Switch to **Run mode** to test your flow. If the flow is ready for use, you can share or publish it.
**Note**
The quality of the generated flow depends on the clarity and specificity of your prompt. You may need to refine your prompt or make manual adjustments to achieve your desired outcome.
## Creating a flow from scratch (blank flow)
Creating a flow from scratch gives you complete control over the design and functionality of your flow.
1. Sign in to the Quick web experience.
1. In the navigation pane, choose **Flows**.
1. Choose **Create Flow**.
1. Create a blank flow.
1. Enter a name and optional description for your flow.
1. In the Flow builder, add and configure the steps you need:
+ Add user input steps to collect text or files from users.
+ Add AI response steps to generate content from chat agents, web search, general knowledge, or image generation.
+ Add data insight steps to get responses from Quick data or dashboards and topics.
+ Add reasoning groups to control how steps run with conditions, loops, or validation.
+ Add action steps to read or write to connected applications.
1. Choose **Save** to save your flow.
For more information about the different components you can use in a flow, see [Terminology and key concepts](terminology-and-key-concepts.md).
## Creating a flow by duplicating an existing flow
Duplicating a flow creates a copy that you can modify for your own purposes.
1. Sign in to the Quick console.
1. In the navigation pane, choose **Flows**.
1. Find the flow you want to duplicate.
1. Choose the ellipsis (⋮) next to the flow name, then choose **Duplicate**.
1. Modify the name and optionally the description for your purposes.
1. Make any necessary modifications to the flow.
1. Choose **Save** to save your flow.
**Note**
When you duplicate a flow, all components, connections, and configurations are copied. However, you may need to reconfigure certain settings, such as authentication for action connectors.
## Creating a flow from an agent conversation
You can create a flow from a conversation with a chat agent. For example, if you have been working through a task like reading an email, augmenting it with internal or web knowledge, and drafting a response, you can convert that conversation into a reusable flow.
1. Sign in to the Quick console.
1. Start a conversation with a chat agent.
1. Work through the task you want to automate.
1. Choose the Quick Flows icon in the chat bar.
1. Choose **Create Flow from conversation**. Amazon Quick Flows interprets the interactions in your conversation and brings you to the Create flow screen with a generated natural language prompt, name, and description.
1. Review and edit the prompt, name, and description as needed.
1. Choose **Generate Flow**.
1. Review the generated flow and make any necessary adjustments.
1. Choose **Save** to save your flow.
## Flow creation requirements
When creating a flow, keep the following requirements in mind:
Flow name
Flow names must be unique within your Quick account and can contain up to 128 characters.
Flow components
A valid flow must have at least one output step or action step.
Reasoning groups
Each reasoning group must have at least one instruction that defines what the AI model should do with the inputs.
Action steps
Action steps require proper authentication and configuration before they can be used in a flow.
## Next steps
After creating a flow, you can:
+ Test your flow to ensure it works as expected.
+ Publish your flow to make it available to users. See [Publishing changes](versioning.md#publishing-your-flow).
+ Share your flow with specific users or groups. See [Sharing flows](sharing-flows.md).
+ Monitor your flow's usage and performance. See [Progress tracker](interacting-with-flows-in-runtime-mode.md#progress-tracker).
# Editing flows
After creating a flow, you can edit and configure it in the Flow editor.
## Accessing the Flow editor
1. Sign in to the Amazon Quick console.
1. In the navigation pane, choose **Flows**.
1. Find the flow you want to edit.
1. Choose the flow tile, or choose the ellipsis (⋮) and select **Open**.
## Configuring step types
Most steps with a prompt field support @ references to include data from previous steps. See each step's configuration for specific limitations.
### Configuring text input steps
In Editor mode, add or select a text input step. In the configuration panel, set the following:
+ **Title**: A name for the step.
+ **Placeholder**: Optional text that appears inside the input field when it is empty. This text is not used for the flow run.
+ **Default Value**: Optional. The value provided if the user doesn't enter an input.
+ **Allow override of default value**: Toggle to let users replace the default value at runtime.
### Configuring file upload steps
In Editor mode, add or select a file upload step. In the configuration panel, set the following:
+ **Title**: A name for the step.
+ **Default file**: Optional. Upload a document, image, or video to use if the user doesn't provide one. You can upload one file per step.
+ **Allow override of default value**: Toggle to let users replace the default file at runtime.
### Configuring general knowledge steps
In Editor mode, add or select a general knowledge step. In the configuration panel, set the following:
+ **Output preference**: Choose one:
+ **Fast responses** — Across image, video, and text inputs.
+ **Versatility and performance** — Balanced capabilities for diverse tasks.
+ **Prompt**: Write the prompt that instructs the model what to generate.
+ **Creativity Level**: Optional. Adjust the slider from low to high to control the randomness of the response.
For more information, see [General knowledge](ai-response-steps.md#general-knowledge-step).
### Configuring Quick data steps
In Editor mode, add or select a Quick data step. In the configuration panel, set the following:
+ **Prompt**: Write the prompt that instructs what content to retrieve.
+ **Link specific resources**: Select spaces and knowledge bases to get insights from. By default, responses are generated from all knowledge sources the user has access to.
### Configuring web search steps
In Editor mode, add or select a web search step. In the configuration panel, set the following:
+ **Prompt**: Write the prompt that instructs what content to search for on the web.
### Configuring research steps
In Editor mode, add or select a research step. In the configuration panel, set the following:
+ **Title**: A name for the step.
+ **Research objective**: Describe what you want to research.
+ **File uploads**: Optional. Upload default files to help guide your research.
+ **Research materials**:
+ **Preferred websites**: Optional. Specify websites or types of websites the agent should prioritize (for example, government websites, academic journals).
+ **Websites to avoid**: Optional. Specify websites or types of websites to exclude (for example, social media, blogs).
+ **Data and apps**: Select all data and apps, or choose specific ones.
### Configuring chat agent steps
In Editor mode, add or select a chat agent step. In the configuration panel, set the following:
+ **Title**: A name for the step.
+ **Chat agent**: Select the agent to use.
+ **Prompt instruction**: Write the prompt that instructs the agent.
+ **Data and apps**: Optionally narrow down the selected data and apps to refine your use case.
+ **Web search**: Toggle to enable or disable web search for the agent.
### Configuring UI Agent steps
In Editor mode, add or select a UI Agent step. In the configuration panel, set the following:
+ **Title**: A name for the step.
+ **UI Agent Instructions**: Write the instructions for the UI agent. Use single, complete URLs for faster, more accurate results.
### Configuring Create Image steps
In Editor mode, add or select a Create Image step. In the configuration panel, set the following:
+ **Prompt**: Describe the image to generate.
+ **Advanced settings**: Configure creativity level, exclude terms, and image seed.
### Configuring Dashboards and topics steps
In Editor mode, add or select a Dashboards and topics step. In the configuration panel, set the following:
+ **Quick Sight source**: Choose from Dashboard or Topic.
+ **Prompt**: Describe the insights you want from your dashboard or topic.
For more information, see [Dashboards and topics](data-insight-steps.md#dashboards-and-topics-step).
### Configuring action steps
In Editor mode, add or select an action step. In the configuration panel, set the following:
+ **Action connector**: Select the connector to use (for example, Salesforce, Jira, Slack).
+ **Action**: Select the specific action to perform.
+ **Prompt**: Write prompt instructions to execute your actions.
For more information, see [Action steps](action-steps-in-flows.md).
### Configuring reasoning groups
In Editor mode, add or select a reasoning group. In the configuration panel, set the following:
+ **Instructions**: Write instructions that tell the model what to do, such as conditions, loops, or validation logic. You can reference the output of a previous step as the input to a loop or conditional statement.
## Prompt writing for steps
When writing prompts for steps, consider the following best practices:
+ Be clear and specific about what you want the model to generate.
+ Provide context to help the model understand the task.
+ Specify the desired format, tone, and style of the output.
+ Use specific language to control the output — for example, "respond in bullet points", "limit the response to three sentences", or "use a formal tone".
+ Use examples to illustrate the expected output when appropriate.
## Adding and removing steps
### Adding steps
1. In the Flow editor, choose **Add step** from the toolbar.
1. Select the type of step you want to add.
1. Drag the step to the desired position in your flow.
1. Configure the step as needed.
### Removing steps
1. Select the step you want to remove.
1. Choose **Delete** from the context menu.
1. Confirm the deletion when prompted.
**Note**
When you remove a step, any @ references to that step in other steps are also removed. You may need to update other steps to maintain your flow.
## Publishing changes
After making changes to your flow, publish them to make them available to users.
1. In the Flow editor, choose **Save** to save your changes.
1. Choose **Publish**.
1. Choose **Publish** to confirm.
When you publish, the changes become immediately available to all users who have access to the flow.
## Updating Flow details
### Updating title and description
1. In the Flow editor, directly edit the title and description in-line.
1. Choose **Save**.
1. Choose **Publish**.
### Viewing the original prompt
If your flow was created using a natural language prompt, you can view the original prompt:
1. In the Flow editor, select the kebab menu in the header.
1. Choose **View prompt**.
**Note**
The original prompt is read-only. To create a new flow based on a modified prompt, use the natural language prompt method.
## Best practices for editing flows
+ Test your changes thoroughly before publishing them to ensure they work as expected.
+ Use clear and descriptive names for steps to make your flow easier to understand and maintain.
+ Write detailed prompts and instructions to get the best results from AI models.
+ Use @ references to create dynamic flows that adapt to user inputs.
+ Consider the user experience when designing the flow of steps.
+ Document your changes in the publication description to maintain a clear history of updates.
Step configurations and available features may change over time. For the latest information, see the [Terminology and key concepts](terminology-and-key-concepts.md) and the [Amazon Quick Flows product page](https://aws.amazon.com/quicksight/q-apps/).
# Versioning
Amazon Quick Flows uses versioning to let you edit a flow without affecting the version that others are using. You can have up to three versions of a flow at any time.
## How versioning works
Your organization's setup determines which version types you see.
### If your organization uses approval review
You work with three version types:
+ **Draft** — Your working copy. You can edit and test your flow without affecting the published version. As you edit, your flow is saved automatically. You can have one draft at a time.
+ **Pending Approval** — A version submitted for review. You can view it and check the approval status, but you cannot edit it. If changes are requested, update your draft and resubmit.
+ **Published** — The approved version that other users can access and run.
For more information about approval reviews, see [Approval review](approval-review.md).
### If your organization doesn't use approval review
You work with two version types:
+ **Draft** — Your working copy. You can edit and test your flow without affecting the published version. As you edit, your flow is saved automatically. You can have one draft at a time.
+ **Published** — The version that other users can access and run. You can publish directly when your flow is ready.
| Version Type | Can Edit? | Visible to Users? | Next Step |
| --- | --- | --- | --- |
| Draft | Yes | No | Publish (or submit for approval) |
| Pending Approval\$1 | No | No | Wait for approval |
| Published | No | Yes | Create new draft for changes |
\$1Pending Approval only applies if your organization uses approval review.
## Publishing changes
Any changes to a flow's title, description, steps, logic, or sharing permissions require publishing. When you publish, the new version replaces the current published version for all users. End users always see only the latest published version.
If your organization uses approval review, publishing submits your flow for review. Once approved, it becomes the new published version. Without approval review, publishing makes changes available immediately.
## Unpublishing flows
You can unpublish a shared flow to remove it from general availability while keeping it accessible to co-owners. Viewers lose access, and the flow returns to draft state. Unpublishing does not require approval review.
# Sharing flows
When you share a flow, you give others permission to view and run it, or to collaborate on it as co-owners.
## Sharing options
You can share a flow in three ways:
+ **With individuals** — Share with specific people using their email addresses.
+ **With groups** — Share with pre-defined groups in your organization (for example, departments or project teams).
+ **With everyone** — Make the flow available to all users in your Amazon Quick Flows instance.
Depending on your organization's settings, sharing may require approval review. For more information, see [Approval review](approval-review.md).
## How to share a flow
1. Open your flow.
1. Choose **Share**.
1. Select how you want to share: with individuals, groups, or everyone.
1. For individuals, enter email addresses. For groups, search for and select the group.
1. Choose the permission level: **Viewer** or **Co-owner**.
1. Choose **Share**.
## Permission levels
### Viewer
Viewers can run the flow and view results, but cannot edit, share, or delete it. Viewers only see the published version.
### Co-owner
Co-owners can edit the flow, share it with others, manage permissions, publish changes, and access draft versions. Co-owners cannot remove the original owner. Only one person can edit at a time.
**Permission reference**
| Permission | Can run? | Can edit? | Can share? | Can delete? |
| --- | --- | --- | --- | --- |
| Viewer | Yes | No | No | No |
| Co-owner | Yes | Yes | Yes | Yes |
# Approval review
When approval review is enabled, users must submit their flows for review before they can be shared. This gives administrators oversight of what flows are available in the organization.
## How approvals work
When approval review is enabled, sharing a flow follows a three-stage cycle:
+ **Submitted** — The flow is waiting for review. You can view it and check the status, but cannot edit it. You can withdraw the request if needed.
+ **Rejected** — The flow was not approved. You can view the reviewer's feedback, update your draft, and resubmit.
+ **Approved** — The flow is shared with your intended audience. Any future changes require a new approval cycle.
## Who can review flows
Users with Amazon Quick Enterprise subscriptions (Author Pro or Admin Pro roles) can review and approve flows. Users with Amazon Quick Professional subscriptions cannot review flows.
Reviewers can:
+ View all submitted flows in the **Pending approval** tab
+ Approve or reject flows individually or in bulk
+ Provide feedback when rejecting a flow
## Submitting a flow for approval
1. Complete and test your flow.
1. Choose **Share** and specify who should have access.
1. Choose **Submit for review**.
1. Monitor the approval status in the **Pending approval** tab.
If your flow is rejected, review the feedback, update your draft, and resubmit.
## Enabling and disabling approval review
Administrators control approval review through the Custom Permissions page.
**When enabled:**
+ All future flow sharing requires approval.
+ Existing shared flows remain accessible.
+ Users can still create and edit flows in draft mode.
**When disabled:**
+ All pending approval requests are automatically rejected.
+ Users can share flows immediately without approval.
+ Previously approved flows remain shared.
## Approval states reference
| State | Can edit? | Visible to users? | Available actions |
| --- | --- | --- | --- |
| Draft | Yes | No | Submit for approval |
| Submitted | No | No | View status, withdraw request |
| Rejected | Yes | No | Update and resubmit |
| Approved | No | Yes | Create new draft for changes |
# Running flows
These topics cover how to run flows in different runtime modes, download outputs, and schedule recurring runs.
# Interacting with flows in runtime mode
When you open a flow, you select **Run mode** to execute it. There are three ways to start a flow.
## Starting a flow with structured input
If the flow has input steps, enter the required information and choose **Start**. The flow runs through each step in sequence and displays the results.
## Starting a flow with conversational runtime
You can supply the flow input (optional) and ask the agent to start the flow through the chat panel. The agent collects any additional inputs, runs the steps, and presents results within the conversation. You can ask follow-up questions or refine outputs as the flow runs.
## Starting a flow from a chat agent
You can run a flow directly from a chat agent conversation:
1. Choose the Flows menu in the chat footer.
1. Select the flow you want to run.
1. Follow the prompts to execute it.
1. When finished, choose **End** to return to your chat agent conversation.
Details from your conversation can automatically populate the text input to your flow. When the flow completes, the output is shared with the calling agent.
## Progress tracker
The progress tracker shows the status of each step as the flow runs.
## Resuming runs from history
You can view and resume previous flow runs from the flow's history. Open the flow and choose the history icon to view previous runs.
# Downloading output in Amazon Quick Flows
After a flow run completes, you can download one or more step outputs as a Word (.docx) or PDF document. You can save files locally or to a shared Quick space for team collaboration. Downloaded files maintain the original formatting, including markdown structure, visual elements, and data tables.
## Download flow outputs
1. In Run mode, choose the download icon (downward arrow) in the top right.
1. Choose **Download as**.
1. Select which outputs to include:
+ Choose **Select All** to include all steps in chronological order.
+ Use checkboxes to select specific steps.
**Note**
User input and file upload steps cannot be downloaded.
1. Select your preferred file format:
+ **Word (docx)** — Editable document format
+ **PDF** — Fixed-layout document format
1. Choose your download destination:
+ **Download** to save the file locally.
+ **Save to a space** to save to a shared space (requires edit access).
# Scheduling your Amazon Quick Flows
Schedules let you automate recurring flow runs without manual intervention. You can schedule flows you have created or flows that have been shared with you. All schedules are private to you and cannot be shared with other users.
Common uses include generating recurring reports, summarizing open items from external services, or preparing daily meeting briefings.
## Creating a schedule
1. Open your flow in Run mode.
1. Choose the scheduling icon.
1. Choose **Create schedule**.
1. Configure your schedule:
+ **Schedule name** — A unique name for the schedule.
+ **Description** — Optional. Details about the schedule's purpose.
+ **Repeat configuration** — Choose from suggestions or configure custom recurrence (daily, weekly, or monthly) with start date, optional end date, and time zone.
1. Choose **Next**.
1. Provide default inputs for each scheduled run.
1. Choose **Next**.
1. Configure action permissions:
+ Turn on **Run with no confirmation** to automatically submit action forms during scheduled runs.
+ Turn off to review and confirm each action manually.
1. Choose **Save**.
**Note**
Action permissions only appear if the flow contains action steps. Automatic form submission is enabled by default for write actions. User confirmation is recommended to help avoid AI prediction errors.
## Managing schedules
You can manage schedules from two places:
+ **From the flow** — Open the flow, choose the scheduling icon, and edit, pause, duplicate, view runs, or delete a schedule.
+ **From the Flows library** — Choose the **My schedules** tab to see all your schedules across all flows with their status, frequency, last run, and associated flow name.
## Schedule history
All manual and scheduled runs appear in the flow run history. You can filter by all runs, scheduled runs, or failed runs. History is retained for 30 days.
Visual indicators help distinguish run types:
+ Schedule run tag for scheduled runs
+ Orange dot for un-viewed completed runs
+ Running tag for in-progress runs
+ Failed tag for failed runs
**Note**
You cannot view an ongoing scheduled run until it completes.
## Notifications
You receive email notifications when:
+ A scheduled run completes successfully (with a link to the results)
+ A run requires your input to continue
+ A run requires authentication for a third-party application
+ A flow with your schedules is unshared from you (schedules are archived)
+ A flow with your schedules is deleted
+ A flow with your schedules is updated with a new version (schedules continue with the updated version)
## Authentication for action connectors
Ensure your authentication for action connectors is current before scheduling. If authentication expires before a scheduled run, you receive an email notification. To verify, run the flow manually or visit the action connector page in Quick to sign in.
# Managing Quick Flows
As an administrator, you can manage how Amazon Quick Flows operates within your organization, including governance policies, user permissions, and feature availability.
## Administrative access requirements
To manage Quick Flows settings, you must have administrator privileges in your Amazon Quick instance. Administrative controls are available through the Amazon Quick console's management interface, where you can configure permissions and manage assets.
## Governance capabilities overview
Quick Flows administrative controls provide multiple layers of governance:
+ **Feature enablement** - Control whether flows are available to users in your account
+ **Approval workflows** - Require administrative approval before flows can be shared with users
+ **Asset management** - Oversee Flow sharing, ownership, and visibility across your organization
# Managing Quick Flows permissions and assets
Administrators manage Quick Flows permissions and assets through two interfaces in the Quick console: the Custom Permissions page and the Asset Management page.
## Custom Permissions page
The Custom Permissions page controls which Quick Flows capabilities and features are available to users.
### Restrict capabilities
+ **Enable or disable flows** — Control whether Quick Flows is available to users in your account. When disabled, all Quick Flows functionality is unavailable, but existing flows are preserved. Quick Flows is enabled by default for new Quick instances.
### Restrict features
+ **Allow creators to share without approval** — When enabled, creators can share flows without review. When disabled, all eligible users can review and approve flow sharing requests.
+ **Enable bedrock model usage in General knowledge step for output refinement** — Controls whether Bedrock models are available for General knowledge steps. Enabled by default.
+ **Enable UI agent to perform browser tasks** (Preview) — Controls whether UI agent steps can be used in flows.
## Asset Management page
The Asset Management page provides oversight of all flows in your organization.
+ **Share flows** — Share flows on behalf of creators, redistribute flows to different groups, or manage sharing permissions across the organization.
+ **Transfer ownership** — Transfer flow ownership between users, such as when a creator leaves the organization.
+ **Unlist flows** — Remove flows from the shared library, making them unavailable to viewers while preserving them for the creator to edit. Unlisted flows can be re-shared once issues are resolved.
# Quick Flows limits
## Flow limits
The following limits apply to flow creation and management in your Quick instance:
+ **Maximum flows per instance:** 10,000 flows
+ **Maximum flows per user:** 100 flows
+ **Maximum steps per flow:** 35 steps
## Input limits
Input limits vary depending on whether you are using General knowledge or Quick data as your data source.
### Text input limits
+ **General knowledge:** 40,000 characters
+ **Quick data:** 15,000 characters
### File upload limits
File upload limits depend on your knowledge source and file type:
**File size limits:**
+ **General knowledge:**
+ Document files: Up to 50 MB (supported file types depend on output preference)
+ Video files: Up to 1 GB
+ Image files: Up to 4.5 MB
+ **Quick data:**
+ Document files: Up to 50 MB (.pdf, .txt, .rtf, .doc, .docx, .ppt, .pptx, .csv, .xls, .xlsx)
+ Image files: Up to 10 MB (.png, .jpg, .jpeg)
**Maximum number of files:**
+ **General knowledge:** Up to 5 files
+ **Quick data:** Up to 20 files
**Character limits for file content:**
+ **General knowledge:**
+ Faster responses: Up to 1M characters total context limit
+ Versatility and Performance: Up to 1M characters total context limit
The total context limit includes characters from output prompts defined at build time and characters from input steps and file upload steps at runtime.
+ **Quick data:** 665,000 characters
## Output limits
Output limits control the size of prompts and generated responses:
+ **Maximum input prompt size:**
+ General knowledge: 5,000 characters
+ Quick data: 5,000 characters
+ **Maximum output characters rendered:**
+ General knowledge: Up to 40,000 characters
+ Quick data: 8,000 characters
## Reasoning group limits
+ **Maximum iterations per reasoning group:** 50 iterations
When referencing the output of a reasoning group, especially looped content, the output is bound by the input character limits of the receiving step. For larger outputs, consider staging the results and querying them in a later step.
## Schedule limits
+ **Region availability**: Schedules in flows are currently supported in US East (N. Virginia), US West (Oregon), and Europe (Ireland)
+ **Maximum schedules per user**: 20 schedules per user
+ **Maximum schedules per instance**: 10,000 schedules
## Regional availability
Certain features have regional limitations:
+ **Image generation:** Currently supported in US East (N. Virginia), US West (Oregon), and Europe (Ireland)