DISCOVER THE FUTURE OF AI AGENTSarrow_forward

Claude Quickstarts

calendar_todayAdded Feb 23, 2026
categoryAgent & Tooling
codeOpen Source
PythonWorkflow AutomationDockerNext.jsMultimodalAI AgentsWeb ApplicationSDKBrowser AutomationAgent & ToolingDeveloper Tools & CodingAutomation, Workflow & RPAProtocol, API & Integration

A collection of official quickstart projects by Anthropic for the Claude API, featuring complete sample applications for customer support, data analysis, computer control, browser automation, and autonomous coding, deployable via Next.js, Docker, and Python.

Claude Quickstarts Project Overview#

Project Summary#

Claude Quickstarts is an open-source monorepo maintained by Anthropic on GitHub, containing multiple independent sub-projects. It provides near-production-ready application templates, not just code snippets.

Core Values:

  • Lower Barrier to Entry: Complete end-to-end code covering frontend, backend, and AI logic
  • Best Practices Demo: Shows how to properly integrate Claude API, handle Tool Use, design prompts, and manage context
  • Multi-Scenario Coverage: Dialog agents, data analysis, desktop automation, browser control, and autonomous coding

Use Cases:

  • Developers needing to quickly build PoCs or MVPs
  • Engineers learning advanced Claude API usage (Tool Use, Computer Use)
  • Architects seeking AI Agent architecture references (ReAct loops, multi-agent collaboration)

Core Sub-Projects#

1. Customer Support Agent#

RAG-based intelligent customer service system.

Key Features:

  • Amazon Bedrock Knowledge Bases integration for enterprise knowledge retrieval
  • Real-time thinking visualization: Displays agent reasoning chain
  • Sentiment detection: Identifies user emotions and triggers appropriate responses
  • Citation support: Shows source document fragments
  • Customizable UI: Built with shadcn/ui, supports dynamic sidebar configuration

Tech Stack: Next.js 14 (App Router) + React + TailwindCSS + shadcn/ui + Anthropic SDK + AWS SDK

Quick Start:

cd customer-support-agent
npm install
echo "ANTHROPIC_API_KEY=your_key" > .env.local
npm run dev
# Visit http://localhost:3000

2. Financial Data Analyst#

Data analysis application with multimodal data understanding and interactive chart generation.

Key Features:

  • Multi-format file parsing: Supports Text, PDF, and Image uploads
  • Interactive visualization: Line charts, bar charts, pie charts, area charts, stacked area charts (Recharts)
  • Streaming response: Low-latency interaction via Next.js Edge Runtime

Applicable Scenarios: Financial data analysis, environmental data analysis, sports performance tracking, social media analysis, education progress tracking, health and fitness monitoring

Tech Stack: Next.js 14 + React + TailwindCSS + Recharts + PDF.js + Anthropic SDK

Quick Start:

cd financial-data-analyst
npm install
echo "ANTHROPIC_API_KEY=your_key" > .env.local
npm run dev
# Visit http://localhost:3000

3. Computer Use Demo#

Demonstrates Claude's ability to control desktop computer environments via natural language.

Key Features:

  • Multi-model support: Claude 4 series (Opus 4.5, Sonnet 4.5, Sonnet 4, Opus 4, Haiku 4.5) and Claude 3.x series
  • Multi-backend support: Claude API, Amazon Bedrock, or Google Vertex AI
  • Containerized isolation: Complete Ubuntu desktop environment in Docker
  • Remote viewing: Real-time screen view via VNC or Web interface

Core Loop: User Prompt → Claude Reasoning → Tool Use → Environment Update → Screenshot → Next Step

Quick Start:

export ANTHROPIC_API_KEY=your_api_key
docker run \
    -e ANTHROPIC_API_KEY=$ANTHROPIC_API_KEY \
    -v $HOME/.anthropic:/home/computeruse/.anthropic \
    -p 5900:5900 -p 8501:8501 -p 6080:6080 -p 8080:8080 \
    -it ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest
# Visit http://localhost:8080

Environment Variables:

  • ANTHROPIC_API_KEY: Claude API key
  • API_PROVIDER: Optional bedrock or vertex
  • WIDTH / HEIGHT: Screen resolution (recommended 1024x768)

4. Browser Tools API Demo#

Playwright-based browser automation reference implementation.

Key Features:

  • DOM interaction: Click, input, scroll, form operations, text extraction
  • Smart targeting: Precise positioning via element ref or coordinates
  • Visual capture: Page screenshots and visual state analysis
  • Page operations: Smart scrolling, page search, navigation control

Architecture:

Docker Container
├── Streamlit Interface (User interaction)
├── Claude API + Browser Tool (Claude controls browser)
├── Playwright + Chromium (Browser automation)
├── XVFB Virtual Display (Virtual display)
└── VNC/NoVNC Server (Visual access)

Quick Start:

cd browser-use-demo
cp .env.example .env
# Edit .env file to add ANTHROPIC_API_KEY
docker-compose up --build
# Visit http://localhost:8080

5. Autonomous Coding Agent#

Dual-agent system based on Claude Agent SDK for autonomous multi-file software development.

Key Features:

  • Dual-agent mode: Initializer Agent (planning) + Coding Agent (execution)
  • Multi-session building: Build complete applications across multiple sessions
  • Progress persistence: Saves development state via feature_list.json and Git commits
  • Sandbox security: OS-level isolation, filesystem restrictions, Bash command whitelist

Workflow:

  1. Initializer analyzes requirements and generates feature list
  2. Coding agent reads feature list, writes code and executes tests item by item
  3. Auto-commits to Git after completing each feature node

Quick Start:

npm install -g @anthropic-ai/claude-code
cd autonomous-coding
pip install -r requirements.txt
export ANTHROPIC_API_KEY='your-api-key-here'
python autonomous_agent_demo.py --project-dir ./my_project

CLI Parameters:

  • --project-dir: Project directory (default: ./autonomous_demo_project)
  • --max-iterations: Maximum agent iterations (default: unlimited)
  • --model: Claude model to use (default: claude-sonnet-4-5-20250929)

Environment Setup#

Prerequisites:

  • Git
  • Node.js (v18+)
  • Python (3.10+)
  • Docker (for Demo projects)

API Key: Get Claude API Key from https://console.anthropic.com

General Installation:

# 1. Clone repository
git clone https://github.com/anthropics/claude-quickstarts.git

# 2. Navigate to specific project
cd claude-quickstarts/[project-name]

# 3. Install dependencies
npm install  # Node.js projects
pip install -r requirements.txt  # Python projects

# 4. Configure environment
echo "ANTHROPIC_API_KEY=your_key" > .env.local

# 5. Run application
npm run dev  # or docker-compose up

Related Projects

View All arrow_forward

STAY UPDATED

Get the latest AI tools and trends delivered straight to your inbox. No spam, just intelligence.

rocket_launch