Darwin
ResearchPricingCase Studies
For Agents
For BusinessOverviewContactFor Marketing TeamsFor Operations TeamsFor Customer Support TeamsFor Sales TeamsFor Finance Teams
For DevelopersOverviewDocsFor AI AssistantsFor MarketplacesFor Shopping AppsFor Productivity AppsFor Finance Apps
CompanyAboutNewsCareersContactSupportInvestorsPrivacy Policy

Loading page…

Darwin
ResearchPricingCase Studies
For Agents
For BusinessOverviewContactFor Marketing TeamsFor Operations TeamsFor Customer Support TeamsFor Sales TeamsFor Finance Teams
For DevelopersOverviewDocsFor AI AssistantsFor MarketplacesFor Shopping AppsFor Productivity AppsFor Finance Apps
CompanyAboutNewsCareersContactSupportInvestorsPrivacy Policy

Darwin Composer

Explore

ResearchAgent Execution

One interface.
For people and agents.

Darwin Composer studies how to turn an agent response into an interface for the task and the caller. It uses DarwinUI components, agent presentation instructions, and the context the caller is permitted to use.

↗Spreadsheet agent · Needs your approval

Export the cleaned spreadsheet

Review the destination before continuing.

Destination
Download
Permission
Owner approval
Interactive interface example using DarwinUI’s DecisionRequest. Decisions update this example only; no file is exported.

On this page

Interfaces for the taskInterfaces that know the contextHow Composer worksBuilt from componentsPeople, agents, and permissionsAcross hostsExplore Composer

Agent work is more than a message. It can involve a choice to review, a task to follow, or a result to inspect. Composer chooses how to present each moment, using DarwinUI components for people and Markdown and structured results for agent callers.

An interface for the next step

A paragraph can describe a decision. A control lets someone make it. Progress should show what has actually happened, and a result should make its details easy to inspect. Try the same task information as text and as an interactive view below.

Clean this spreadsheet and let me review the export.

Text description

The spreadsheet has been cleaned. The next step is to export a CSV. Review the destination and approve the export before continuing.

DarwinUI interaction

↗Spreadsheet agent · Needs your approval

Review the export

Choose whether to continue with this destination.

Format
CSV
Destination
Download
Interactive examples using DarwinUI’s DecisionRequest, WorkProgress, and FieldList. All data and task states are illustrative. Controls change this page only; no file is exported and no agent is contacted.

The interface follows the state supplied by the application. An approval changes a decision, a progress view reflects recorded steps, and a result keeps its summary and exceptions together. The examples are intentionally small enough to understand and use in place.

An interface that adapts to you

Personalization can shape the whole experience: which inputs are already filled in, what needs clarification, how results are presented, and what a useful next step looks like. A shopping agent might start with a preferred size and budget; a research agent might organize findings around the questions and sources that matter to you. Those choices should remain visible and editable.

Composer uses the caller’s identity, message, target agent, operation, and returned response. Agent presentation instructions and the available DarwinUI components guide the interface. Runtime supplies verified operation state and enforces access and approvals.

We study how this adaptation continues inside a thread: preserving corrections, keeping earlier results accessible, and adjusting the next interaction as the task changes. For an AI participant, the same context should inform structured inputs and outputs. Remembering preferences belongs to the account and application; the interface makes their effects understandable and correctable.

How Composer works

Darwin Composer uses the caller’s intent and the agent response to generate an interface.ParticipantsPeople & agentsChoose operationsreview decisionsDarwin ComposerAdaptive interfaceDarwinUI componentsMarkdown & dataYour applicationRuntimeAuthorize actionsreturn application stateInputs & decisionsProgress & resultsAction intentsPresentation stateIdentity, permissions, and execution belong to your applicationDarwin Composer uses the caller’s intent and the agent response to generate an interface.ParticipantsPeople & agentsChoose operationsreview decisionsInputs & decisionsProgress & resultsDarwin ComposerAdaptive interfaceDarwinUI componentsMarkdown & dataAction intentsPresentation stateYour applicationRuntimeAuthorize actionsreturn application stateIdentity, permissions, and executionbelong to your application
Runtime executes authorized work and returns its state. Composer decides how to present that state to the caller.

A person needs clear controls and a way to review a decision. An AI needs a structured account of the operation and its current state. These are different presentations of the same underlying interaction.

DarwinUI brings agents, operations, progress, approvals, and results into a reusable interface model. The example above switches between a human view and an illustrative structured view without changing the decision being represented.

Build with the same components

The example uses the real DarwinUI DecisionRequest component. An application supplies the state and handles a decision; the component presents that state and the available controls.

React
1import { DecisionRequest } from "@darwinui/react";2 3<DecisionRequest4  title="Export the cleaned spreadsheet"5  agent="Spreadsheet agent"6  description="Review the destination before continuing."7  details={[{ label: "Destination", value: "Download" }]}8  decision={decision}9  onDecision={setDecision}10  approveLabel="Approve export"11  declineLabel="Keep for review"12/>
The component and props are present in the DarwinUI library. Your application owns authorization and the action behind the callback.

Shared work needs explicit authority

A shared thread is useful only when participants can understand who is directing the work and who may approve a sensitive action. Our interface research keeps those responsibilities visible: members can direct the AI; the owner approves sensitive actions.

Interface concernWhat remains explicit
InputsWhat the operation needs, which values were filled from context, and what can be changed.
ProgressThe current task state, interruptions, and results that have actually arrived.
ApprovalWho can decide, what the decision covers, and whether it has been given.
ResultsThe returned artifact or information, distinct from a request being accepted.
Presentation helps a participant understand an operation. Runtime permission checks remain authoritative.

A consistent model across different hosts

React interfaces, embedded MCP experiences, and API consumers have different presentation constraints. We study which interaction semantics should remain consistent across them, and which controls should adapt to the host.

A reusable view does not make a host capable of executing every action. Supported operations, session access, approvals, and returned results still need to be checked in the actual host.

Explore Composer

DarwinUI provides the private component library behind Composer, including reusable agent, thread, operation, decision, progress, and result components.

Our research asks how these pieces can compose into useful interfaces while keeping the operation, evidence, and authority understandable. For personalization, we ask whether people repeat themselves less, can correct an assumption easily, and keep control as the interface adapts. We also evaluate accessibility, state continuity, and each interaction in its host.

Explore Composer.

See how adaptive presentation fits alongside agent execution.

Explore Darwin Runtime

Explore more research

View all

Darwin Builder

An AI coding system built for creating agents: their operations, implementation, and interface library.

Explore

Darwin Compiler

From agent definitions to validated execution and interfaces.

Explore

Darwin Runtime

Coordinating agents from intent to outcome.

Explore
Darwin

ProductsNavigatorAgents By DarwinPricing
ResourcesFor DevelopersFor BusinessCase Studies
CompanyNewsAboutCareers
LegalTerms of ServicePrivacy PolicyAcceptable Use Policy

© 2026 Darwin Technologies, Inc.