ClaudevsKiro

A focused deep-dive into Anthropic Claude and Amazon Kiro — two Claude-powered platforms with radically different purposes, architectures, and use cases.

Published: May 2026
Claude: 5 Tiers
Kiro: 5 Tiers
Sections: 8
Status: ● Live Data
Two Platforms, One Engine
Claude powers both products — but what you get from each is fundamentally different.

Claude.ai and Kiro both run on Anthropic's Claude models, but they represent opposite ends of the AI product spectrum. Claude.ai is a general-purpose conversational AI assistant — designed for research, writing, analysis, and broad knowledge work across industries. Kiro is a specialized agentic developer IDE built by Amazon/AWS, designed exclusively for software engineering workflows with deep codebase integration, spec-driven development, and CI/CD automation. Choosing between them isn't an "either/or" — many developers use both, with Claude.ai for exploration and writing, and Kiro for implementation.

🤖
Claude
Anthropic · claude.ai · General-Purpose AI

A conversational AI assistant accessible via web, mobile, and desktop. Claude handles writing, research, analysis, coding assistance, document processing, and agentic tasks. Available to individuals, teams, and enterprises at multiple price points. Governed by Anthropic's Constitutional AI safety framework and Public Benefit Corporation legal structure.

Best for: General knowledge work, research, writing, multi-domain tasks
Kiro
Amazon/AWS · kiro.dev · Agentic Dev IDE

An agentic AI IDE and CLI built by Amazon, powered by Claude (and other models) via Amazon Bedrock. Kiro introduces spec-driven development — turning natural language prompts into structured requirements, architecture designs, and executable tasks across entire codebases. Supports VS Code extensions, Open VSX plugins, and deep AWS integration including GovCloud.

Best for: Software engineers, spec-first development, large codebases, CI/CD automation
Scope note: All pricing, privacy scores, and data practices represent researcher estimates based on official published policies as of May 2026. Verify all figures against official provider documentation before making procurement decisions.
The Model Relationship
Kiro is not a competitor to Claude — it's a product built on top of Claude via Amazon Bedrock.
Architecture Overview
Anthropic
Model Creator
Claude Opus · Sonnet · Haiku
Constitutional AI · Safety Research
API via
Amazon Bedrock
Kiro (AWS)
Application Layer
IDE · CLI · Spec Engine
Hooks · Agent Orchestration

Kiro routes model requests through Amazon Bedrock using cross-region inference. Claude models (plus DeepSeek, MiniMax, GLM, Qwen) are all available. Kiro's "Auto" mode selects the best model per task. Enterprise Kiro data never leaves AWS infrastructure and is never stored.

Kiro's model advantage: Kiro offers Claude Opus 4.7 (Experimental, launched April 16, 2026) — a model not yet available in Claude.ai's consumer interface. Kiro also offers DeepSeek, MiniMax, GLM-5, and Qwen3 Coder at significantly lower credit costs (0.05x–0.25x Auto baseline), giving developers budget-conscious options unavailable in Claude.ai.
MCP: NPX & UVX Servers
Kiro's MCP layer connects to external tools, services, and data sources. Two runtime environments cover the entire ecosystem: NPX for JavaScript/Node.js tools and UVX for Python tools.

Model Context Protocol (MCP) is the open standard that lets Kiro reach beyond its built-in capabilities and interact with external services, APIs, databases, and documentation. Kiro supports both local MCP servers (run as child processes on your machine) and remote MCP servers (HTTPS endpoints). All configuration lives in a JSON file — either workspace-level at .kiro/settings/mcp.json or globally at ~/.kiro/settings/mcp.json.

📦
NPX
Node Package Execute · JavaScript/TypeScript ecosystem

What it is: NPX runs Node.js packages directly without a global install. Kiro uses command: "npx" with args: ["-y", "@package/name"] — the -y flag auto-confirms the download, making first-run completely silent.

Prerequisite

Node.js must be installed. Download from nodejs.org. Verify with node --version.

NPX MCP Servers Available

  • Amazon Devices Builder Tools
  • Azure resource management
  • Chrome DevTools (browser automation)
  • Context7 (live library documentation)
  • Docker container management
  • Dynatrace observability
  • GitHub repository access
  • Brave Search web search
  • Filesystem access tools
  • Postgres/SQLite/MySQL database tools
// .kiro/settings/mcp.json — NPX example { "mcpServers": { "brave-search": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-bravesearch"], "env": { "BRAVE_API_KEY": "${BRAVE_API_KEY}" } }, "docker": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-docker"] }, "chrome-devtools": { "command": "npx", "args": ["-y", "chrome-devtools-mcp@latest"] } } }
🐍
UVX
UV Tool Runner · Python ecosystem

What it is: UVX is the tool-runner component of UV — an extremely fast Python package manager from Astral.sh (written in Rust). It runs Python MCP servers in isolated environments without requiring a global Python setup or virtual environment management. Kiro uses command: "uvx" with the PyPI package name directly.

Prerequisite

UV must be installed. One command: curl -LsSf https://astral.sh/uv/install.sh | sh (macOS/Linux) or winget install --id=astral-sh.uv (Windows). Verify with uvx --version.

UVX MCP Servers Available

  • AWS Documentation (search & recommendations)
  • CrowdStrike Falcon security analysis
  • Python data science tools
  • AWS CLI wrapper tools
  • Custom Python MCP servers (awslabs.* packages)
  • Any Python package published to PyPI
// .kiro/settings/mcp.json — UVX example { "mcpServers": { "aws-docs": { "command": "uvx", "args": ["awslabs.aws-documentation-mcp-server@latest"], "env": { "FASTMCP_LOG_LEVEL": "ERROR" } }, "crowdstrike": { "command": "uvx", "args": ["falcon-mcp"], "env": { "FALCON_CLIENT_ID": "${FALCON_CLIENT_ID}", "FALCON_CLIENT_SECRET": "${FALCON_CLIENT_SECRET}" } } } }
NPX vs UVX — Capability Comparison (Researcher Scores, 0–10)
Relative scores across key dimensions for each runtime. NPX wins on ecosystem breadth; UVX wins on isolation, speed of cold-starts after first download, and Python data tooling.
NPX (Node.js) UVX (Python)

Remote MCP Servers (No Runtime Required)

Some MCP servers are hosted remotely and accessible via HTTPS — no Node or Python required. These use the "url" property instead of "command". Examples include Amplitude (analytics), Datadog (observability), and Databricks (data platform). Remote servers are ideal for enterprise SaaS integrations with authentication headers.

// Remote MCP — connect to a hosted HTTPS endpoint { "mcpServers": { "amplitude": { "url": "https://mcp.amplitude.com/mcp", "disabled": false, "autoApprove": [] }, "databricks": { "url": "${DATABRICKS_SQL_MCP_URL}", "headers": { "Authorization": "Bearer ${DATABRICKS_ACCESS_TOKEN}" } } } }

autoApprove

List specific tool names that run without a confirmation prompt. Use "*" to auto-approve all tools in a server. Best used for read-only tools (documentation, search) — never for tools that modify data.

Example: "autoApprove": ["search_docs", "get_recommendations"]

disabledTools

Exclude specific tools from appearing to the agent, even if the server offers them. Useful for servers that expose dangerous tools (e.g., a database server's DELETE tools) while keeping safer ones available.

Example: "disabledTools": ["delete_table", "drop_database"]

Scope levels

Workspace config (.kiro/settings/mcp.json) applies to one project — check this into version control for team consistency. User config (~/.kiro/settings/mcp.json) applies globally across all workspaces — for personal tools like web search or your note-taking system.

Security rule: Always use environment variable references (${VAR_NAME}) rather than hardcoding API keys. Never commit config files with credentials. Use autoApprove only for read-only tools. Review a server's full tool list before connecting.
Agent Skills: Kiro vs Claude Code
Both platforms have a "skills" concept — but they work fundamentally differently in scope, portability, and activation mechanism.

Kiro Agent Skills follow the open agentskills.io standard — portable packages that any compatible AI IDE can use. Claude Code has a comparable (but simpler) concept through custom slash commands and CLAUDE.md instructions. Understanding the differences helps you choose the right tool for each type of reusable workflow.

Dimension Kiro Agent Skills Claude Code Equivalent
StandardOpen standard (agentskills.io) — portable across compatible IDEsProprietary to Claude Code — custom commands in .claude/commands/
File formatFolder with SKILL.md + optional scripts, references, assets directoriesMarkdown .md files in .claude/commands/; referenced via slash commands
ActivationAutomatic (matches description to request) OR explicit /skill-name slash commandExplicit only: /command-name slash command or direct mention
Progressive loadingAt startup only names & descriptions load; full instructions load on demand — saves contextFull content loaded when invoked; no lazy loading
ScopeWorkspace (.kiro/skills/) OR global (~/.kiro/skills/)Workspace (.claude/commands/) — no native global scope
Executable scriptsYes — scripts/ subfolder; Kiro runs scripts for deterministic tasks (validation, generation)No — instructions only; Claude Code itself executes commands
Sharing / importImport from GitHub URL or local folder — one click in IDE panelManual copy of .md files; no formal import mechanism
Template variablesVia scripts and references directoryYes — $ARGUMENTS placeholder captures inline text after command
Team distributionVersion-controlled in repo (.kiro/skills/) — automatic for all team membersVersion-controlled in repo (.claude/commands/) — same benefit
Conflict resolutionWorkspace skill overrides same-named global skillN/A — no global scope

Kiro Skill — Anatomy

# .kiro/skills/pr-review/SKILL.md --- name: pr-review description: Review pull requests for code quality, security issues, and test coverage. Use when reviewing PRs or preparing code for review. --- ## Review process 1. Check for security vulnerabilities (OWASP Top 10) 2. Verify all error paths are handled 3. Confirm test coverage for new logic 4. Review naming consistency with codebase 5. Confirm no secrets or credentials in diff # References (loaded only when needed) # references/security-checklist.md # references/team-conventions.md

Skill auto-activates when you ask Kiro to "review this PR" or "check this code for issues." Or invoke directly: /pr-review

Claude Code Command — Anatomy

# .claude/commands/pr-review.md Review the following pull request for: - Security vulnerabilities (OWASP Top 10) - Unhandled error paths - Test coverage for new logic - Naming consistency with codebase conventions - Hardcoded secrets or credentials in the diff PR details or diff to review: $ARGUMENTS # Usage: /pr-review [paste diff or PR URL here]

Invoked explicitly: /pr-review — the $ARGUMENTS variable captures any text typed after the command.

Agent Skills Feature Comparison — Kiro vs Claude Code
Scored 0–10. Kiro wins on portability and automation; Claude Code wins on template variables and simplicity.
Kiro Agent Skills Claude Code Commands
When to use which: Use Kiro Agent Skills for reusable workflows you want to share across IDEs or with your team via GitHub. Use Claude Code commands for simpler, single-project prompts where template variables ($ARGUMENTS) are the main need. Both can co-exist — many teams version-control both.
Steering: Kiro vs Claude Code CLAUDE.md
Both platforms let you give the AI persistent, project-aware context — but Kiro's steering is dramatically more granular, with four inclusion modes and file-pattern targeting.
Dimension Kiro Steering Claude Code CLAUDE.md
File locationWorkspace: .kiro/steering/*.md · Global: ~/.kiro/steering/*.mdWorkspace: CLAUDE.md (root) or any subdirectory · Global: ~/.claude/CLAUDE.md
Multiple filesYes — unlimited steering files, each focused on one domainYes — CLAUDE.md in any directory; Claude Code discovers all of them
Inclusion modes4 modes: always, auto, fileMatch, manualAlways included when in scope directory — no conditional loading
File-pattern targetingYes — fileMatchPattern: "components/**/*.tsx" loads only when editing matching filesPartial — directory-level CLAUDE.md files apply to that subtree
Auto-activation by descriptionYes — inclusion: auto mode matches context to a description, like skillsNo
Manual / on-demandYes — inclusion: manual + reference by #filename in chat or /filename slash commandNo explicit mechanism — all in-scope files always load
Live file referencesYes — #[[file:api/openapi.yaml]] embeds live file contentIndirect — Claude Code reads files you mention, but no embedded reference syntax
Foundation generatorBuilt-in: auto-generates product.md, tech.md, structure.md from codebase analysisNo — write CLAUDE.md manually
AGENTS.md supportYes — Kiro reads AGENTS.md files at root or ~/.kiro/steering/Yes — CLAUDE.md is effectively the same standard
Team distributionMDM/Group Policy push to ~/.kiro/steering/ for org-wide defaultsGit-tracked CLAUDE.md; no push mechanism for global
Refinement assistBuilt-in "Refine" button — Kiro helps improve your steering fileNo

The Four Kiro Steering Inclusion Modes

always

Loaded into every interaction, no conditions. Use for core standards: tech stack, security policies, naming conventions. These form the baseline of Kiro's understanding of your project.

--- inclusion: always ---

fileMatch

Loads only when you're working on files matching a glob pattern. Keeps context focused — your React component conventions only load when editing *.tsx files.

--- inclusion: fileMatch fileMatchPattern: "src/**/*.tsx" ---

auto

Kiro reads the description and loads the file when your request matches it — like a steering version of Agent Skills' progressive disclosure. Great for specialised context that's heavy but occasional.

--- inclusion: auto name: api-design description: REST API design and endpoint conventions ---

manual

On-demand only. Reference with #steering-filename in chat or /steering-filename slash command. Best for rarely-needed but detailed documents like migration guides or runbooks.

--- inclusion: manual --- # invoked via: /migration-guide

Kiro Steering — Full Example

# .kiro/steering/api-standards.md --- inclusion: fileMatch fileMatchPattern: ["app/api/**/*", "**/*.controller.ts"] --- ## REST API Standards Endpoints: Use nouns, not verbs. Good: GET /users/:id Bad: GET /getUser/:id Error format: Always return structured errors: { "error": { "code": "NOT_FOUND", "message": "..." } } Authentication: All endpoints require JWT except /auth/* and /health # Reference live OpenAPI spec #[[file:api/openapi.yaml]]

Claude Code CLAUDE.md — Equivalent

# app/api/CLAUDE.md # (placed in subdirectory — loads for all files under app/api/) ## REST API Standards Endpoints: Use nouns, not verbs. Good: GET /users/:id Bad: GET /getUser/:id Error format: Always return structured errors: { "error": { "code": "NOT_FOUND", "message": "..." } } Authentication: All endpoints require JWT except /auth/* and /health # See also: api/openapi.yaml for full spec

Claude Code discovers this file automatically for all interactions within app/api/. No frontmatter needed — simpler, but always-on with no pattern matching.

Bottom line: If you only need "always apply these rules to this directory," Claude Code's CLAUDE.md is simpler and equally effective. If you need context to activate conditionally (only for TypeScript, only for API files, only on demand), Kiro's steering system is meaningfully more powerful — and its live file references keep steering in sync with your actual codebase automatically.
Hallucination in Kiro: Causes & Mitigation
Agentic AI tools like Kiro can "hallucinate" — generating plausible-sounding but incorrect code, API calls, imports, or library usage. Kiro has structural features that dramatically reduce this risk when used correctly.

Hallucination in a coding agent is different from hallucination in a chat assistant. In a chat assistant, you might get a wrong fact. In an agentic IDE, you might get code that compiles but uses a function that doesn't exist in your version of a library, references a file path that doesn't exist, or silently breaks a behavior three files away from the change. The good news is that Kiro's spec-driven architecture was designed specifically to counter the most common causes of agent drift and fabrication.

Common Causes of Hallucination in Kiro

1. Stale or missing context

The agent makes assumptions about files it hasn't read. Most common in long sessions or when working across many files without spec guidance. The agent fills gaps with plausible-but-wrong details — wrong function signatures, invented imports, non-existent API endpoints.

2. Vague prompts

Underspecified requests leave the model to invent constraints. "Add authentication" is ambiguous: which library? which token format? which endpoints? The model invents answers to unasked questions, and those inventions may not match your actual stack.

3. Session length drift

In very long Autopilot sessions, models (especially Haiku and Sonnet) lose track of earlier context and may contradict prior decisions. Opus maintains focus better but is not immune. The longer the session, the higher the drift risk.

4. Wrong model for task complexity

Using Haiku for complex multi-file refactoring, or using Auto when the task requires deep architectural reasoning, significantly increases hallucination probability. Model selection is a key risk control.

5. No steering / no specs

Without steering files, the agent has no ground truth about your conventions. Without specs, there are no acceptance criteria to validate against. Both create conditions where the model fills gaps with training data patterns rather than your actual codebase patterns.

Kiro's Built-In Mitigation Mechanisms

1. Specs with acceptance criteria

Specs force the model to write down what "done" looks like before writing code. Each task maps to specific requirements. Kiro can validate implementations against acceptance criteria — a verifiable check that hallucinated code will fail.

Impact: High — most effective single mitigation for complex features

2. Checkpoint system

Kiro creates restore points before major changes. Any drift can be identified by reviewing diffs and reverted instantly to a known-good state. Treating checkpoints as mini code reviews catches hallucinations before they compound.

Impact: Medium — reactive but low friction to use

3. Supervised mode

Switching from Autopilot to Supervised for sensitive sections forces a human review at each step. The agent suggests changes but cannot commit them — you see the diff and decide. Near-zero hallucination risk in supervised mode at the cost of speed.

Impact: Very High — but slower; use for critical production changes

4. Steering with live file references

Using #[[file:path]] in steering embeds the actual current content of key files (OpenAPI specs, config templates, example components) directly into the model's context. The model sees real code, not its memory of code.

Impact: High — grounds the model in actual codebase reality

5. Test hooks (auto-verify after change)

Configure a hook to run your test suite automatically after every file save. Hallucinated code that compiles but misbehaves gets caught immediately — before the next change compounds the error.

Impact: High — continuous verification loop
Hallucination Risk by Model & Task Complexity — Relative Scores (Higher = More Risk)
Researcher estimates. All models improve substantially with specs and steering. Opus 4.7's self-verification capability is reflected in its lowest scores. Use higher-numbered Opus for complex or high-risk tasks.
Opus 4.7 Opus 4.6 Sonnet 4.6 Haiku 4.5 Auto

Developer Mitigation Checklist

Before Starting

  • Create a spec for any task spanning >2 files
  • Verify steering files are current and cover your stack
  • Add live #[[file:]] references for key schemas/specs
  • Choose Opus for complex reasoning; Auto for routine tasks
  • Start a fresh session (don't carry over unrelated prior context)

During Development

  • Review code diffs at each task boundary — not just at the end
  • Run tests via hook after every meaningful change
  • Switch to Supervised mode for security-critical code
  • Use checkpoints before large refactors
  • If the model seems confused, interrupt and use "View all changes" to assess drift

After Implementation

  • Validate each task against its spec acceptance criteria
  • Run full test suite — not just the new tests
  • Check imports and dependencies actually exist in package.json / requirements.txt
  • Review any new function signatures against actual library docs
  • Revert to checkpoint if more than ~20% of changes look wrong
The spec principle: The single most effective anti-hallucination practice is using specs for anything non-trivial. When the model writes requirements and acceptance criteria before touching code, it constrains its own solution space and creates verifiable targets. Hallucination thrives on ambiguity — specs eliminate ambiguity by design.
Kiro Full Model Catalog
All models available in Kiro as of May 2026 — covering Claude, open-weight, and budget alternatives. Cost is relative to Auto (1.0× baseline).

Kiro's model selection is one of its key differentiators vs Claude.ai — you're not locked into a single model family. The right model for a task depends on complexity, budget, and whether you need Claude's Constitutional AI guarantees or are comfortable with open-weight alternatives for non-sensitive work.

Model Performance Profile — Kiro Models Compared (0–10)
Across intelligence, coding strength, reasoning depth, speed, and credit efficiency. Grouped by model family for clarity.
Opus 4.7 (Exp.) Opus 4.6 Sonnet 4.6 Haiku 4.5 Auto Qwen3 / Budget
Model Family Context Cost (×Auto) Free Status Ideal For Self-Verify
Claude Opus 4.7 Anthropic 1M 2.2× Experimental Hardest agentic tasks, verifying its own outputs, high-res vision ✓ Strong
Claude Opus 4.6 Anthropic 1M 2.2× Active Large codebases, planning ahead, debugging complex issues ✓ Good
Claude Opus 4.5 Anthropic 200K 2.2× Active Maximum reasoning, multi-system tradeoff analysis ~ Moderate
Claude Sonnet 4.6 Anthropic 1M 1.3× Active Multi-agent pipelines, near-Opus quality, long sessions ~ Moderate
Claude Sonnet 4.5 Anthropic 200K 1.3× Active Best agentic coding default, extended autonomous runs ~ Moderate
Claude Sonnet 4.0 Anthropic 200K 1.3× Active Consistent baseline, no routing layers, predictable behavior
Auto ⭐ Kiro Router 1.0× baseline Active Default for most work — picks the best model per task automatically Varies
Claude Haiku 4.5 Anthropic 200K 0.4× Active Quick iterations, sub-agent orchestration, credit conservation
MiniMax M2.5 Open Weight 200K 0.25× Experimental Near-Opus coding at low cost, full dev lifecycle
DeepSeek 3.2 Open Weight 128K 0.25× Experimental Agentic workflows, multi-step reasoning at low cost
GLM-5 Open Weight 200K 0.5× Experimental Repo-scale work, long-horizon agentic tasks across large codebases
MiniMax M2.1 Open Weight 200K 0.15× Experimental Multilingual: Rust, Go, C++, Kotlin, TypeScript, UI generation
Qwen3 Coder Next Open Weight 256K 0.05× Experimental Most cost-efficient; long coding sessions; strong error recovery
Model selection guide: Start with Auto. Switch to Opus 4.6 or 4.7 when you hit a wall on complex multi-file work or need self-correction. Use Haiku or Qwen3 for quick iterations where burning fewer credits matters.
Experimental models: These are production-capable but may have limited region availability or change without notice. For CI/CD pipelines, use Active models to avoid unexpected behavior changes.
Privacy note: Open-weight models (DeepSeek, MiniMax, GLM, Qwen) are hosted on AWS Bedrock infrastructure, not on the model providers' own servers. Enterprise data protections still apply.
Anthropic Claude — Tier Summary
Five subscription tiers designed for individuals to large enterprises.
Free
$0
Per month

Sonnet 4.6 + Haiku 4.5. Limited usage windows. No Opus, no Projects, no Claude Code.

Pro
$20/mo
$17/mo annual

5× usage vs Free. Full Opus 4.6 + Sonnet 4.6 + Haiku 4.5. Claude Code, Projects, 200K context.

Max
$100–200/mo
5× or 20× vs Pro

Max 5× or 20× usage. 1M context in Claude Code. For heavy daily users and power workflows.

Team
$25–100/seat
Min. 5 seats

Standard ($25/seat) or Premium ($100 w/ Claude Code). SSO, admin, M365/Slack. Training OFF contractually.

Enterprise
Custom
Contact sales

500K context, HIPAA, SCIM, audit logs, RBAC, custom retention. SOC 2 Type II. Training OFF contractually.

Privacy standout: Claude is the only major AI platform where training is OFF by default at the consumer level (opt-in, not opt-out). Team and Enterprise plans disable training contractually — not just as a settings toggle.
Kiro (Amazon/AWS) — Tier Summary
Kiro uses a credit-based pricing model. All paid plans support pay-per-use overages at $0.04/credit. GovCloud pricing is ~20% higher.
Free
$0
Per month
50 credits/mo

Auto + Sonnet 4.0 + Sonnet 4.5 + some budget models. 500 bonus credits for 30 days on first signup. No Opus.

Pro
$20/mo
+ $0.04/credit overage
1,000 credits/mo

Full model access incl. Opus 4.6. IP Indemnity. Autopilot + Supervised mode. All Kiro features.

Pro+
$40/mo
+ $0.04/credit overage
2,000 credits/mo

Double the Pro credits. Same features. Best for heavy spec work or frequent Opus usage. IP Indemnity included.

Power
$200/mo
+ $0.04/credit overage
10,000 credits/mo

Heavy team leaders, CI/CD pipeline automation, frequent Opus 4.7 usage. Equivalent to Claude Max in price.

Enterprise
Custom
Contact AWS sales

Data NOT stored. SAML/SCIM SSO. Customer-managed KMS keys. GovCloud support. Org management. User activity reports.

Understanding Kiro's Credit System

Unlike Claude's session-based usage model, Kiro charges per prompt using a credit system. Credits deplete differently depending on which model you use and the complexity of the task. The Auto model (1.0× baseline) is the cheapest way to get Sonnet-class results. Opus models cost 2.2× more credits per task.

Credit Cost Multipliers by Model (Relative to Auto = 1.0×)
Lower multiplier = fewer credits consumed per task. Budget models like Qwen3 Coder Next (0.05×) are 44× cheaper than Opus per credit.
Credit math example: On the Kiro Pro plan (1,000 credits/mo), using Auto exclusively you'd get ~1,000 meaningful tasks. Switch to Opus 4.7 (2.2× cost) and that drops to ~455 tasks. Switch to Qwen3 Coder Next (0.05× cost) and your 1,000 credits would handle ~20,000 tasks — effectively unlimited for most developers.
All Available Models in Kiro
Kiro offers the full Claude lineup plus several open-weight alternatives via Amazon Bedrock. This is a key differentiator vs Claude.ai which only offers Claude models.
Model Context Credit Cost Free Pro+ Status Best Use
Claude Opus 4.71M2.2×ExperimentalHardest agentic tasks, self-verifying code review
Claude Opus 4.61M2.2×ActiveDeep reasoning, large codebases, debugging
Claude Opus 4.5200K2.2×ActiveMulti-system complexity, ambiguity handling
Claude Sonnet 4.61M1.3×ActiveNear-Opus quality, multi-agent pipelines
Claude Sonnet 4.5200K1.3×ActiveAgentic coding default, extended autonomous runs
Auto (recommended)1.0× baselineActiveRoutes to best model per task automatically
Claude Haiku 4.5200K0.4×ActiveQuick iterations, sub-agent orchestration
DeepSeek 3.2128K0.25×ExperimentalAgentic workflows, multi-step reasoning at low cost
MiniMax M2.5200K0.25×ExperimentalNear-Opus coding results at low cost
GLM-5200K0.5×ExperimentalRepo-scale work, long-horizon agentic tasks
Qwen3 Coder Next256K0.05×ExperimentalLong coding sessions, most cost-effective available
Opus 4.7 availability: As of April 16, 2026 Kiro is one of the only platforms with access to Claude Opus 4.7 (Experimental). It is currently restricted to AWS IAM Identity Center users, with broader rollout pending. This model is not yet available in Claude.ai's consumer interface.
Feature Comparison Matrix
Side-by-side capabilities for Claude.ai and Kiro across the dimensions that matter most to developers and teams.
Feature / Capability Claude.ai Kiro
Primary Use CaseGeneral AI assistant — chat, research, writing, coding helpAgentic software engineering — spec, build, deploy
Free Tier✓ Sonnet 4.6 + Haiku 4.5✓ 50 credits + 500 bonus on signup
Max Context Window1M tokens (Claude Code)1M tokens (Opus 4.6/4.7, Sonnet 4.6) + 256K (Qwen)
Opus 4.7 Access✗ Not yet available✓ Experimental (AWS IDC users)
Multi-Model Choice✗ Claude models only✓ Claude + DeepSeek + MiniMax + GLM + Qwen
Spec-Driven Development✗ No structured spec system✓ Native (Requirements → Design → Tasks)
Local Codebase Agent✓ Claude Code (CLI, all OS)✓ Kiro IDE + CLI (all OS)
Agent Hooks (event-triggered)✗ No✓ On file save, custom triggers
VS Code Extension Compatibility~ Via extensions✓ Full Open VSX + VS Code settings import
CI/CD Pipeline Integration~ Via API✓ Native CLI pipeline integration
Web Search / Research✓ Deep Research (built-in)~ URL context fetch (limited)
Projects / Memory✓ Projects w/ Files, Memory, Instructions~ Steering files (project-level context)
Artifact Generation✓ HTML, React, SVG, code, charts~ Code + docs via spec output
Image/Multimodal Input✓ Upload images, documents, code✓ Drop UI designs or whiteboard photos
MCP App Connections✓ Google Drive, Gmail, Jira, Miro, Slack, etc.✓ MCP + remote MCP support
Autopilot / Auto-accept Mode✓ Claude Code auto-accept✓ Autopilot mode (default on)
Enterprise SSO✓ SAML/SCIM (Team+)✓ SAML/SCIM via AWS IAM Identity Center, Okta, Entra
GovCloud / Government Support✗ No✓ AWS GovCloud (US) — ~20% price premium
Training Off by Default✓ Opt-in only (consumer)~ Opt-out available (consumer); Auto-off (enterprise)
Enterprise Data Storage✓ Configurable retention✓ Enterprise data NOT stored at all
IP Indemnity✗ Not offered✓ Pro, Pro+, Power subscribers
HIPAA Ready✓ Enterprise tier✓ AWS compliance programs
Desktop App✓ macOS + Windows✓ macOS + Windows + Linux
Mobile App✓ iOS + Android✗ No mobile app
Image Generation✗ No✗ No
Overage Billing✗ No overage, usage resets✓ $0.04/credit overage (paid plans)
Student Pricing✗ No formal student plan✓ Students program available
Cost Analysis — Pricing by Tier
Monthly USD pricing for all tiers across both platforms. Note: Kiro's $40 Pro+ tier has no Claude equivalent — the gap between $20 Pro and $200 Power is bridged for mid-volume developers.

Both platforms share a $20 entry point and a $200 power-user tier, but Kiro adds a unique $40 Pro+ tier filling the gap for mid-volume developers. Claude has the advantage of a $100 Max tier. The deeper difference is the billing model: Claude uses session/rolling-window limits while Kiro uses a discrete credit system with pay-per-use overage — giving developers more granular spend control.

Side-by-Side Pricing Comparison — All Tiers
Monthly USD. Enterprise shown as lowest common estimate for comparison. Claude Max split into 5× ($100) and 20× ($200) variants.
Claude (Anthropic) Kiro (Amazon/AWS)
TierClaudeKiroKey Difference
Free $0 · Sonnet 4.6 + Haiku 4.5 · Limited sessions $0 · 50 credits/mo + 500 bonus (first 30 days) · Auto + Sonnet Kiro's bonus trial credits make it more generous at first; Claude's free tier has no credit cap
First Pay Pro · $20/mo ($17 annual) · 5× usage · Opus access Pro · $20/mo · 1,000 credits · Full model access + IP Indemnity Claude includes Claude Code; Kiro includes IP Indemnity. Both include their flagship agentic coding tool.
Mid-Tier Max 5× · $100/mo · 5× usage vs Pro Pro+ · $40/mo · 2,000 credits
Power · $200/mo · 10,000 credits
Kiro's $40 tier fills a gap. Claude's Max 20× ($200) = Kiro Power at same price — both target the same heavy-user segment
Team Standard · $25/seat · Premium · $100/seat (w/ Code) Enterprise plan · Assign individual tier per user · Centralized billing Claude has dedicated team pricing; Kiro's "team" is enterprise-only with per-user plan assignment
Enterprise Custom sales · 500K context · HIPAA · SOC 2 Custom sales · Data NOT stored · GovCloud · Customer-managed KMS Kiro enterprise data is never stored; Claude enterprise has configurable retention. Kiro adds GovCloud.
Value insight: At the identical $20/month price, Claude Pro offers Claude Code, Projects with persistent memory/files/instructions, and 200K context. Kiro Pro offers 1,000 credits (usable across any model), IP Indemnity, VS Code-compatible IDE, spec-driven development, and CLI pipeline integration. They complement rather than replace each other.
Consumer Privacy Scores by Tier
Overall privacy score (0–10, higher = better). Derived from training opt-out mechanism, retention policy, third-party sharing, human review access, and user control options.
Methodology: Composite of: training default (30%), retention duration (25%), third-party sharing (20%), human review access (15%), user controls (10%). Researcher estimates from published policies as of May 2026.
First Pay Tier — Privacy Score
Claude Pro (opt-in training, 30-day retention if opted out) vs Kiro Pro (opt-out available, data stored in US-East with service improvement use). Higher = better.

Claude Pro — Score: 8/10

  • Training: OFF by default — must opt-in explicitly
  • Retention: 30 days if opted out; up to 5 years if opted in
  • Human review: Possible for flagged content; minimal otherwise
  • Third-party sharing: Minimal — no advertising ecosystem
  • User controls: Clear opt-in/opt-out toggle in Settings

Kiro Pro — Score: 6/10

  • Training: ON by default (service improvement); opt-out available in Settings
  • Retention: Data stored in US-East (N. Virginia) indefinitely
  • Human review: Possible per AWS support access policies
  • Third-party sharing: Routed through Amazon Bedrock; multiple model providers
  • User controls: Opt-out available; telemetry also separately controllable
Second Pay Tier — Privacy Score
Claude Max vs Kiro Pro+. Both maintain the same privacy posture as their first-pay tiers — no material privacy changes at the second tier for either platform.
Team / Enterprise Tier — Privacy Score
Claude Team/Enterprise vs Kiro Enterprise. Enterprise tiers show the most dramatic improvement for both platforms. Kiro Enterprise data is never stored; Claude Enterprise training is disabled contractually.

Claude Enterprise — Score: 10/10

  • Training: OFF contractually (cannot be changed unilaterally)
  • Retention: Custom configurable, down to 30 days
  • Human review: Restricted by contractual DPA
  • Third-party sharing: DPA governs all data handling
  • Compliance: SOC 2 Type II, HIPAA, GDPR DPA

Kiro Enterprise — Score: 9/10

  • Training: Data NOT stored at all → cannot be used for training
  • Retention: Zero — enterprise data never persisted
  • Human review: Not applicable — no data stored
  • Third-party sharing: AWS shared responsibility model applies
  • Compliance: AWS compliance programs, GovCloud option, customer-managed KMS
Notable: Kiro Enterprise's "data not stored" policy is arguably stronger than Claude Enterprise's configurable retention — there's nothing to breach if data is never written to disk. However, Claude Enterprise's contractual DPA provides explicit legal accountability that Kiro's shared responsibility model leaves partially to the customer.
Data Retention Risk
Worst-case data retention in months. Indefinite = 60. Lower = lower risk. Color: green ≤3mo, amber ≤12mo, orange ≤36mo, red = indefinite.
Data Retention Risk — All Tiers Compared
Each tier shown for both platforms. Consumer individual tiers have the highest retention risk. Enterprise Kiro scores 0 — data is never stored at all.
Low (≤3 mo) Medium (≤12 mo) High (≤36 mo) Indefinite (60)
Key finding: Both Claude and Kiro consumer tiers store data indefinitely by default (60 months for chart purposes). Claude's opt-in model means opting out reduces retention to 30 days. Kiro's retention is harder to characterize because it's stored in US-East and used for service improvement by default unless you opt out. Kiro Enterprise is the unique standout: data is never stored (0 months), making it the most privacy-forward enterprise option analyzed in this report.
Training Data Default
0 = never trains by default (green), 0.5 = trains by default but easy opt-out (amber), 1 = trains by default (red).
Training Data Default — All Tiers Compared
Claude's opt-in model gives it a consistent score of 0 across consumer tiers. Kiro scores 0.5 at consumer level (trains by default, opt-out available) and 0 at enterprise (data not stored).
CLAUDE — SAFE BY DEFAULT (ALL TIERS)

Claude is the only major consumer AI that requires explicit opt-in for training. Free, Pro, and Max users all start with training disabled. Team and Enterprise disable it contractually. This is a key competitive differentiator for privacy-conscious users and regulated industries.

KIRO — OPT-OUT REQUIRED (CONSUMER TIERS)

Kiro Free, Pro, Pro+, and Power users have their data used for service improvement by default. Opt-out is available in Settings (IDE) or Preferences (CLI) and is reasonably accessible. Note: if you have an Amazon Q Developer Pro subscription, Kiro automatically excludes your data from service improvement — another path to privacy.

Multi-Dimensional Privacy Profiles
Radar charts across five risk dimensions. Smaller filled area = lower risk. 0 = safest, 10 = highest risk.
Consumer Tiers — Privacy Radar Comparison
Claude Pro/Max (gold) vs Kiro Pro/Pro+/Power (blue). Claude's opt-in training model gives it a dramatic advantage on training risk. Click legend to toggle.
Enterprise Tiers — Privacy Radar Comparison
Claude Enterprise (gold) vs Kiro Enterprise (blue). Both show dramatically smaller polygons vs consumer tiers. Kiro's zero-storage policy shows on Retention Risk.
Regulatory Action History (2023–2026)
Significant regulatory actions, enforcement proceedings, and major legal events. Corporate-level risk affects all tiers — higher tiers provide contractual protections but not immunity.

Kiro benefits from being a new-to-market product (2025) with Amazon/AWS as its parent entity. AWS carries its own regulatory track record — largely positive — and provides Kiro with compliance infrastructure that would take years to build independently. Anthropic's Claude has minimal direct regulatory history (2 events) thanks largely to its PBC structure and Constitutional AI approach. Kiro, as an AWS product, inherits Amazon's broader compliance posture but also its regulatory surface area.

Regulatory Actions 2023–2026 — Platform Comparison
Anthropic (direct actions) vs Kiro/AWS (AI-specific actions only — general AWS regulatory history excluded). Note: AWS regulatory events related to non-AI services are excluded for fair comparison.

Claude / Anthropic — 2 Events

  • Dec 2025 — State AG Letter: Included in National Association of AGs letter regarding "delusional outputs" and safety safeguards.
  • Apr 2026 — Music Lyrics Copyright Case: Anthropic argues in US case that Claude transforms lyrics rather than reproducing them verbatim. Outcome pending.
  • Notably absent: No GDPR enforcement, no FTC action, no national bans. Anthropic's PBC structure and limited advertising model have kept it largely clear of privacy enforcement.

Kiro / Amazon AWS — ~1 AI-Specific Event

  • Ongoing EU AI Act compliance: AWS/Amazon is subject to EU AI Act GPAI provisions for Kiro as a general-purpose AI interface. Amazon signed EU Code of Practice.
  • AWS parent benefits: Kiro inherits AWS SOC 2, ISO 27001, FedRAMP, HIPAA BAA, and GDPR DPA infrastructure — compliance assets that would take a startup years to acquire.
  • GovCloud moat: Kiro's AWS GovCloud availability gives it access to regulated US government markets that no other AI coding tool has achieved.
Regulatory outlook: Both platforms have notably low regulatory exposure compared to larger AI incumbents (OpenAI/7 events, Google/5 events). Kiro's AWS backing means it benefits from a compliance infrastructure built over 20 years — a significant moat for enterprise adoption in regulated industries.
Kiro: Spec-Driven Development
Kiro's signature capability — turning a natural language idea into structured requirements, architecture, and executable task plans before writing a single line of code.

Spec-driven development is what separates Kiro from every other AI coding tool. Rather than jumping straight into code generation ("vibe coding"), Kiro first creates a structured specification — a formal artifact that documents intent, constraints, architecture decisions, and implementation tasks. This solves the core problem of AI coding tools: they misinterpret context and produce code that diverges from what you actually need as complexity grows.

The Three-Phase Spec Workflow

Phase 1 — Requirements
requirements.md

Your natural language prompt is converted into formal user stories with acceptance criteria in EARS notation (Easy Approach to Requirements Syntax). Every "I want to..." becomes a numbered, testable requirement with explicit constraints.

Example: "The system shall allow users to export reports as PDF when clicking the export button"
Phase 2 — Design
design.md

Kiro analyzes your existing codebase and generates a technical architecture doc: component design, sequence diagrams, data flow, error handling strategy, and testing approach — all specific to your tech stack and existing patterns.

Includes: system architecture, API contracts, database schema changes, security considerations
Phase 3 — Tasks
tasks.md

A discrete, sequenced implementation plan — ordered by dependencies, each task mapped to specific requirements. Real-time status updates as Kiro implements. Run tasks individually or all at once in Autopilot mode.

Tasks are trackable, reversible (checkpoint system), and iterable

Feature Specs

For building new functionality. Two workflow variants:

  • Requirements-First: Start by defining what the feature must do, then design the implementation
  • Design-First: Start with architecture (e.g., you already know the design), then generate requirements retroactively

Best for: new APIs, new UI components, new business logic modules

Bugfix Specs

For systematic debugging with regression prevention:

  • Describes current behavior vs expected behavior vs what should remain unchanged
  • Kiro identifies root cause, designs targeted fix, plans validation tests
  • Prevents the common AI bug where fixing one thing breaks three others

Best for: production bugs, regressions, complex multi-file issues

When to use Specs vs Vibe (chat): Use Specs for complex features requiring structured planning, team collaboration, or documentation. Use Vibe (direct chat) for quick exploratory coding, prototyping, or single-file changes. Specs pay off when implementation spans multiple files and the cost of misinterpretation is high.
Kiro: Agent Hooks & Automation
Hooks let you delegate repeating tasks to AI agents that trigger automatically on events — without needing to prompt manually.

What Are Hooks?

Hooks are pre-defined AI agents that trigger on specific events in your development environment. You configure them once and they run in the background — silently handling recurring tasks that would otherwise require manual prompting.

Common Hook Use Cases

  • On file save: Auto-generate or update unit tests for changed functions
  • On file save: Refresh inline documentation to match current code
  • On commit: Generate a meaningful git commit message from the diff
  • On new file creation: Add standard headers, copyright notices, or boilerplate
  • On test failure: Automatically analyze error and suggest fix
  • On spec task complete: Verify implementation against acceptance criteria

Autopilot vs Supervised Mode

Autopilot Mode (default)

  • Executes multiple steps without per-step approval
  • Makes architectural decisions based on requirements
  • Can create, modify, delete files, and run commands
  • Can be interrupted at any time to regain control
  • All changes visible via "View all changes" in Chat panel

Supervised Mode

  • Suggests actions but waits for user confirmation at each step
  • Asks clarifying questions when requirements are ambiguous
  • Full review of each file change before committing
  • Best for: sensitive codebases, production changes, learning
  • Reverts available at any point via checkpoint system
Trusted commands: Configure exactly which shell commands Kiro can run without confirmation. Use wildcard matching (e.g., "npm *" trusts all npm commands). Universal trust ("*") is available but use with extreme caution.
Kiro: IDE, CLI & Integration
Kiro's dual-surface architecture — a full IDE for interactive work and a CLI for terminal and pipeline automation.
🖥️ Kiro IDE
Download from kiro.dev/downloads · macOS, Windows, Linux

Core IDE Capabilities

  • Built on Code OSS (VS Code base) — familiar environment
  • Import VS Code settings, themes, and Open VSX plugins
  • Multimodal chat — paste UI designs or whiteboard photos
  • Spec panel: create/manage specs, track task status
  • Live code diffs — approve, step through, or edit each change
  • Checkpoint system — revert to any prior state
  • Intelligent error diagnostics — interprets syntax/type/semantic errors
  • Native Git commit message generation from source control pane

Steering Files

Configure how Kiro agents behave per-project or globally. Add coding standards, preferred tools, architectural patterns, and domain-specific context. Similar in function to Claude Project Instructions, but stored as versioned markdown files in your repo.

⚡ Kiro CLI
curl -fsSL https://cli.kiro.dev/install | bash

CLI Capabilities

  • Works in any terminal on macOS, Linux, or Windows
  • Interactive loop — build features, trace bugs, suggest fixes
  • SSH remote support — run against remote servers
  • Custom agents for CI/CD pipeline automation
  • Autocomplete for 500+ popular CLIs
  • Bash, zsh, fish shell support

Pipeline Integration

Trigger Kiro CLI agents within CI/CD pipelines. Automate code review, test generation, documentation updates, and security scanning on push/PR events. This is where Kiro diverges most dramatically from Claude.ai — it can be a fully automated team member in your delivery pipeline.

# Install Kiro CLI curl -fsSL https://cli.kiro.dev/install | bash # Launch in project directory cd /path/to/your/project kiro # Create a spec from terminal kiro spec "Add OAuth2 login with Google" # Run in CI/CD pipeline kiro --non-interactive "Review PR for security issues"
MCP Support: Kiro supports native MCP (Model Context Protocol) integration — the same protocol used by Claude.ai for app connections. Connect Kiro to databases, documentation sites, cloud APIs, and internal tools. Kiro also supports remote MCP — connecting to MCP servers over the network rather than only locally, which Claude.ai does not currently support.
Claude.ai Core Abilities
A walkthrough of the capabilities available in Claude.ai that go beyond what Kiro offers — general intelligence, Projects, research, and design.

Basic Chat Capabilities

  • Long-form analysis up to 200K tokens (1M in Claude Code)
  • Upload PDFs, Word docs, CSVs, images, code files
  • Deep research mode with cited web search
  • Multi-document synthesis and comparison
  • Code generation, debugging, refactoring, SQL
  • Architecture and system design consultation
  • Artifacts: interactive HTML/React, SVG, Mermaid, code files

Projects System

  • Files: Persistent knowledge base — upload SOPs, docs, policies once, reference always
  • Instructions: System prompt for the project — define Claude's role, behavior, tone, constraints
  • Memory: Auto-generated facts about you/your work; reviewable and editable
  • Share: Share entire project context with team members (Team/Enterprise)

Available: Pro, Max, Team, Enterprise

Design Capabilities (Artifacts)

  • Prototypes: Interactive HTML/React/JS apps rendered in-browser
  • Design Systems: Color palettes, component specs, CSS token systems
  • Slide Decks: HTML presentations or PowerPoint (.pptx) via computer use
  • From Template: Adapts from existing UI patterns and component libraries
  • Diagrams: Mermaid ERDs, flowcharts, SVG illustrations, D3/Chart.js visualizations
🔌 MCP App Connections

Claude supports OAuth-authenticated connections to third-party services: Google Drive, Gmail, Google Calendar, Atlassian (Jira/Confluence), Miro, Slack, and more via the MCP registry. Claude asks before using any connected tool — no automatic data access. All connections require explicit user authentication.

⚙️ Skills (Reusable Instruction Sets)

Skills are portable, reusable instruction templates that apply Claude's behavior across multiple Projects and conversations. Create skills for brand voice, coding style, specific analysis frameworks, or domain personas. Team/Enterprise plans can share skills org-wide for consistent AI behavior at scale — comparable to Kiro's steering files but operating at the platform rather than project level.

Claude Model Summary
The three Claude model families available in Claude.ai as of May 2026, plus Adaptive Thinking. (Note: Opus 4.7 is available in Kiro but not yet in Claude.ai.)
Opus 4.6
Flagship · Most Capable
Pro+
Intelligence
9.6
Coding
9.4
Reasoning
9.6
Writing
9.8
Speed
5.5

Best For

  • Complex multi-step reasoning
  • Long-form analysis and research synthesis
  • Advanced code architecture
  • Tasks where quality > speed
API: $5/M in · $25/M out · Context: 200K (1M in Code)
RECOMMENDED
Sonnet 4.6
Balanced · Best Value
All Plans
Intelligence
8.8
Coding
9.0
Reasoning
8.6
Writing
8.7
Speed
7.8

Best For

  • Daily work tasks (emails, docs, analysis)
  • Software development and code review
  • Research assistance and summarization
  • 90%+ of use cases at 40% of Opus cost
API: $3/M in · $15/M out · Context: 200K (1M in Code)
Haiku 4.5
Fastest · Most Affordable
All Plans
Intelligence
7.2
Coding
7.4
Reasoning
6.8
Writing
7.5
Speed
9.6

Best For

  • High-volume, latency-sensitive applications
  • Classification and routing tasks
  • Customer service first responses
  • Sub-agent orchestration within pipelines
API: $0.80/M in · $4/M out · Context: 200K
🧠 Adaptive Thinking (Extended Thinking Toggle)
Available on Sonnet 4.6 and Opus 4.6 · Pro plans and above

When enabled, Claude generates an internal reasoning step before responding — a "scratch pad" where it works through the problem, considers alternatives, checks its own reasoning, and may revise multiple times before committing to a final answer. This significantly improves performance on complex math, code, and multi-constraint analysis.

Enable When

  • Complex math or logic problems
  • Subtle bugs that require tracing root causes
  • Legal/medical/technical analysis with conflicting evidence
  • Multi-constraint optimization problems
  • Research with ambiguous or contradictory sources

Trade-offs

  • Slower response (extra thinking time)
  • More tokens consumed (billed at output rates)
  • Overkill for simple factual queries
  • Toggle on/off per-conversation or via API
  • Thinking token budget configurable via API
Instructions Template
A practical template for writing effective Claude Project Instructions — applicable also as the foundation for Kiro steering files.

Well-written Instructions are the highest-leverage customization in Claude.ai. For Kiro, the equivalent is a steering file — a markdown document checked into your project that tells Kiro agents how to operate. The structure below works for both: paste it into Claude's Project Instructions field, or save it as .kiro/steering.md in your project root.

## Role & Identity You are [ROLE TITLE], a specialized assistant for [ORGANIZATION/TEAM NAME]. Your primary purpose is [CORE MISSION IN ONE SENTENCE]. You have deep expertise in [DOMAIN 1], [DOMAIN 2], and [DOMAIN 3]. ## Audience You are speaking with [AUDIENCE — e.g., "senior software engineers with 5+ years of experience"]. [CALIBRATION — e.g., "Do not over-explain basic concepts. Use technical terminology freely."] ## Operating Context - Company/Project: [NAME] - Industry: [INDUSTRY] - Tech stack: [LANGUAGES / FRAMEWORKS / TOOLS] - Key stakeholders: [WHO DECISIONS GO TO] - Compliance requirements: [HIPAA / SOC2 / GDPR / etc. if applicable] ## Behavioral Guidelines ALWAYS: - [BEHAVIOR 1 — e.g., "Write TypeScript with strict typing — no 'any' types"] - [BEHAVIOR 2 — e.g., "Flag uncertainty explicitly — say 'I'm not certain, but...'"] - [BEHAVIOR 3 — e.g., "Include error handling in all code examples"] - [BEHAVIOR 4 — e.g., "Write tests when generating new functions (Jest / pytest)"] NEVER: - [CONSTRAINT 1 — e.g., "Do not use deprecated APIs or packages"] - [CONSTRAINT 2 — e.g., "Do not store secrets in code or committed environment files"] - [CONSTRAINT 3 — e.g., "Do not introduce new dependencies without noting tradeoffs"] - [CONSTRAINT 4 — e.g., "Do not make commitments about timelines or pricing"] ## Output Format Preferences Default: [prose / bullet points / code-first / structured report] Code output: Code first → brief explanation → edge cases → test example Analysis: Context → findings → recommendation → next steps Length: [brief ~150 words / standard / comprehensive with full explanations] ## Domain Knowledge [KEY CONTEXT about your codebase, architecture, team conventions] Example: "We use microservices on AWS (ECS). Services communicate via SQS. Frontend: Next.js 14 (App Router). Backend: FastAPI (Python 3.12). DB: PostgreSQL 16, Redis. We follow trunk-based development. All PRs require 2 reviews. We use Conventional Commits." ## Tone & Persona Tone: [professional / direct / collaborative / technical] Response style: [concise and direct / comprehensive / Socratic] ## Edge Cases & Escalation If asked about [SENSITIVE TOPIC]: [SPECIFIC RESPONSE] If request is out of scope: [REDIRECT — e.g., "This is outside my domain. Suggest consulting [TEAM]."] Priority rule: If instructions conflict with user request, [PRIORITY — e.g., "follow instructions unless user explicitly overrides with a reason"]

Example: Senior Dev Assistant (Claude Project)

You are a Principal Engineer assistant for the platform team. Stack: TypeScript/React, Python/FastAPI, PostgreSQL, Redis, AWS. ALWAYS: - Strict TypeScript — no 'any' types - 2-space indent, single quotes, no semicolons - Error handling in every code example - Jest tests for new TS functions, pytest for Python - Prefer composition over inheritance NEVER: - Deprecated APIs or packages - Secrets in committed code - New dependencies without tradeoff notes Output: Code first → design decisions → edge cases → test example

Example: Kiro Steering File (.kiro/steering.md)

# Project Steering — Acme API Platform ## Stack - Backend: Python 3.12 / FastAPI / SQLAlchemy - DB: PostgreSQL 16 with Alembic migrations - Auth: JWT + OAuth2 (Google, GitHub) - Infra: AWS ECS + RDS + ElastiCache ## Coding Standards - Type hints on all functions (mypy strict) - Pydantic v2 for all request/response models - 100% test coverage required for new endpoints - Always use async def for DB operations ## Never - Raw SQL queries — use SQLAlchemy ORM - Synchronous I/O in async context - Hardcoded credentials or magic strings ## Preferred patterns - Repository pattern for data access - Service layer for business logic - Dependency injection via FastAPI Depends

Common Role Archetypes — Quick Reference

Technical Roles

  • Principal Software Engineer
  • DevOps / Platform Engineer
  • Data Scientist / ML Engineer
  • Security Engineer (AppSec)
  • Technical Writer
  • QA / Test Automation Engineer

Business Roles

  • Business Analyst
  • Product Manager
  • Strategy Consultant
  • Financial Analyst
  • Market Research Analyst
  • Executive Communications Lead

Kiro Steering Examples

  • Mono-repo navigator with workspace rules
  • Security-first reviewer (OWASP focus)
  • API-first designer (OpenAPI specs)
  • Test coverage enforcer
  • Migration specialist (legacy to modern)
  • CI/CD pipeline assistant

Claude vs Kiro Comparison Report · May 2026 · Research reference document

All pricing, scores, and data practices are researcher estimates based on publicly available information as of May 2026. Verify against official documentation before procurement decisions. Sources: kiro.dev, claude.ai/pricing, docs.anthropic.com, AWS shared responsibility model documentation.

Note: Kiro is an Amazon/AWS product powered by Anthropic Claude via Amazon Bedrock. It is not an Anthropic product. Model availability and pricing in Kiro are governed by AWS service terms, not Anthropic's direct terms.