Loading image...Kiro

Product

  • About Kiro
  • IDE
  • CLI
  • Web
  • Mobile
  • Pricing
  • Downloads

For

  • Enterprise
  • Startups
  • Students

Community

  • Overview
  • Ambassadors
  • Discord
  • Events
  • Powers
  • Shop
  • Showcase

Resources

  • Docs
  • Blog
  • Changelog
  • FAQs
  • Report a bug
  • Suggest an idea
  • Billing support

Social

Site TermsLicenseResponsible AI PolicyLegalPrivacy PolicyCookie Preferences
Loading image...Kiro
  • CLI
  • IDE
  • Web
  • Mobile
  • Enterprise
  • Pricing
  • Docs
SIGN INDOWNLOADS
Loading image...Kiro

Get Started

InstallationAuthenticationYour first project

Models

OverviewAvailable modelsReasoning effort

Features

How Kiro works
Specs
Feature Specs
Bugfix Specs
Quick Spec
Plan mode
Analyze Requirements
Correctness
Best practices
Steering
Hooks
MCP
Permissions
Custom agents
Agent Skills
Powers
CompactionKiroignoreCheckpoints and rewind
Built-in tools
Configuration scopes

IDE 1.x

What's new in 1.0
Setup & First Run
Editor
Chat
Experimental
Troubleshooting0.x reference

CLI

What's new in 3.0
Setup & First Run
Terminal UI
Chat
Headless modeACPAuto complete
Experimental
2.x reference

Web - Preview

Setup & First RunIdentity Center
Connect your repositories
Working with the agent
Autonomous modeAutomations
Sandbox

Mobile - Preview

Overview

Commands and Reference

CLI commandsSlash commandsBuilt-in toolsExit codesSettingsIDE keyboard shortcuts

Billing

OverviewManaging your subscriptionUpgrading your planDowngrading your planCancelling your planPurchasing add-on creditsManaging your paymentsManaging usage notificationsManaging your taxesContacting billing supportDeleting your accountRelated questions

Enterprise

ConceptsOnboarding quickstart
Connecting your identity provider
Subscribe your teamManage subscriptions
Governance
Monitor and track
SettingsManaged updatesBillingIAMSupported regions

Privacy and Security

OverviewData protectionCode referencesCompliance validationInfrastructure securityIAM permissionsFirewalls, proxies, and data perimetersVPC endpoints (AWS PrivateLink)

Guides

Overview
Language support
Learn by playing

Migration

Migrating from Q DeveloperMigrating from VSCodeUpgrading from Q CLI
  1. Docs
  2. Features
  3. Specs
  4. Plan mode

Plan mode


Plan mode helps you think before you act. It guides you through structured requirements gathering and codebase exploration, then produces an implementation plan that hands off to execution. Unlike full Feature Specs, Plan mode doesn't generate formal requirements/design/tasks documents - it's faster and more conversational while still providing structured output.

CapabilityIDECLIWebMobile
Plan mode✓✓——

When to use Plan mode

  • Complex tasks where you want to think through the approach before writing code
  • Multi-file changes where understanding existing architecture matters
  • Situations where you'd normally say "let me understand this first, then we'll implement"

For well-defined features that need formal requirements and design documents, use Feature Specs instead. For quick, well-understood changes, skip planning and work directly with the default agent.

How to invoke

Choose Plan from the workflow options on the "Let's build" screen when starting a new session. To switch mid-session, click the agent name in the bottom bar of the chat input and select Plan from the agent picker.

The agent name in the chat input shows which agent is active.

Workflow

1. Requirements gathering

The planner asks structured questions to refine your initial idea:

bash
[plan] > I want to build a todo app I understand you want to build a todo app. Let me help you plan this. [1]: What platform should this target? a. Web Application b. Mobile App c. CLI Tool d. Other [2]: What's the primary use case? a. Personal Task Management b. Team Collaboration c. Project Management d. Other (Answer with "1=a, 2=b" or provide your own answers)

You can answer individual questions, provide free-form responses, or skip questions entirely.

2. Research and analysis

The planner explores your codebase in read-only mode:

  • Reads files and uses code intelligence to understand existing architecture
  • Uses grep and glob to find relevant patterns
  • Researches technologies via web search
  • Adapts the plan to your existing conventions

3. Implementation plan

Creates a step-by-step plan with clear objectives:

bash
**Implementation Plan - Todo CLI Command** **Problem Statement:** Add todo management to existing application. **Requirements:** - CRUD operations for todos - Local SQLite storage - Priority and due date support **Task Breakdown:** Task 1: Create database schema and models - Define Todo struct with required fields - Create database migration - Demo: Can create and query todos Task 2: Implement command structure - Add subcommand with add/list/complete operations - Demo: CLI accepts todo commands Task 3: Add advanced features - Implement due dates and priority sorting - Demo: Complete system with all features

4. Handoff to execution

Once you approve the plan, execution begins automatically:

  1. Plan mode exits and the agent transitions to execution mode
  2. The complete plan is passed to the execution agent as context
  3. Implementation begins immediately with the approved plan as the guide
Info

In the CLI, automatic plan execution requires the v3 engine. In classic mode, you are prompted to confirm the handoff before execution begins. If a plan doesn't transfer to execution, make sure you explicitly approved it when asked.

Read-only access

Plan mode intentionally cannot modify your project. This keeps the focus on thinking and planning:

OperationAvailable
File reading✓
Code intelligence (symbols, definitions)✓
Search (grep, glob)✓
Web search✓
File writing—
Command execution—
MCP tools—

Plan mode vs Specs

Plan modeFeature Specs
OutputConversational plan with task breakdownFormal requirements.md, design.md, tasks.md
ProcessInteractive Q&A → plan → handoffRequirements → design → tasks with approval gates
SpeedMinutesLonger (multi-phase with reviews)
Best forMedium complexity, explore-then-buildHigh complexity, team collaboration, audit trail
ArtifactsPlan lives in conversation contextPersisted in .kiro/specs/

Tips

  • Use Plan mode for tasks that need 15–60 minutes of implementation - long enough to benefit from planning, short enough that formal specs would be overkill
  • Engage with the structured questions - better input produces better plans
  • Let the planner explore your codebase - it finds patterns you might miss
  • Review the plan before approving handoff - iterate until it matches your expectations
Page updated: August 4, 2026
Quick Spec
Analyze Requirements