/new-app

---

🚨 EXECUTION NOTICE FOR CLAUDE

When you invoke this command via SlashCommand, the system returns THESE INSTRUCTIONS below.

YOU are the executor. This is NOT an autonomous subprocess.

• ✅ The phases below are YOUR execution checklist
• ✅ YOU must run each phase immediately using tools (Bash, Read, Write, Edit, TodoWrite)
• ✅ Complete ALL phases before considering this command done
• ❌ DON't wait for "the command to complete" - YOU complete it by executing the phases
• ❌ DON't treat this as status output - it IS your instruction set

Immediately after SlashCommand returns, start executing Phase 0, then Phase 1, etc.

See @CLAUDE.md section "SlashCommand Execution - YOU Are The Executor" for detailed explanation.

---

Available Skills

This commands has access to the following skills from the claude-agent-sdk plugin:

• fastmcp-integration: Examples and patterns for integrating FastMCP Cloud servers with Claude Agent SDK using HTTP transport
• sdk-config-validator: Validates Claude Agent SDK configuration files, environment setup, dependencies, and project structure

To use a skill:

!{skill skill-name}

Use skills when you need:

• Domain-specific templates and examples
• Validation scripts and automation
• Best practices and patterns
• Configuration generators

Skills provide pre-built resources to accelerate your work.

---

Security Requirements

CRITICAL: All generated files must follow security rules:

@docs/security/SECURITY-RULES.md

Key requirements:

• Never hardcode API keys or secrets
• Use placeholders: your_service_key_here
• Protect .env files with .gitignore
• Create .env.example with placeholders only
• Document key acquisition for users

Arguments: $ARGUMENTS

Goal: Create and setup a new Claude Agent SDK application with TypeScript or Python support

Core Principles:

• Ask before acting - gather language preference and requirements first
• Use latest SDK versions - check npm/PyPI for current releases
• Verify code works - run type checking before completing

Phase 1: Discovery Goal: Gather project requirements and preferences

Actions:

• Parse $ARGUMENTS for project name (if provided)
• Use AskUserQuestion to gather:
  • Language preference: TypeScript or Python?
  • Project name (if not in $ARGUMENTS)
  • Agent type/purpose
  • Starting point preference (minimal, basic, or specific example)
  • Package manager preference (npm/yarn/pnpm for TS, pip/poetry for Python)

Phase 2: Analysis Goal: Load SDK documentation and determine setup approach

Actions:

• Read SDK documentation: ~/.claude/plugins/marketplaces/ai-dev-marketplace/plugins/claude-agent-sdk/docs/sdk-documentation.md
• Check for latest SDK versions
• Identify required dependencies based on language choice
• Determine project structure needs

Phase 3: Planning Goal: Design the project setup approach

Actions:

• Outline project structure:
  • TypeScript: package.json, tsconfig.json, src/index.ts, .env.example
  • Python: requirements.txt, main.py, .env.example
• Plan SDK installation command
• Identify starter code to generate
• Confirm approach with user

Phase 4: Implementation Goal: Execute project setup with specialized agent

Actions:

Task(description="Setup Claude Agent SDK project", subagent_type="claude-agent-sdk:claude-agent-setup", prompt="You are the claude-agent-setup agent. Create a new Claude Agent SDK project for $ARGUMENTS.

Context from user discovery:

• Language: [TypeScript or Python from Phase 1]
• Project name: $ARGUMENTS
• Agent type: [Purpose from Phase 1]
• Package manager: [Preference from Phase 1]

Create the complete project:

• Project directory structure
• Initialize package manager (npm/pip)
• Install Claude Agent SDK with latest version
• Generate starter code with proper SDK usage
• Create .env.example with API key placeholder
• Add .gitignore for security
• Create README.md with setup instructions

Expected output: Fully initialized project ready to run")

Phase 5: Verification Goal: Validate the setup is correct

Actions:

Based on language choice from Phase 1:

If TypeScript:

Task(description="Verify TypeScript setup", subagent_type="claude-agent-sdk:claude-agent-verifier-ts", prompt="Verify the TypeScript Claude Agent SDK setup at $ARGUMENTS.

Check:

• Package.json has correct SDK dependency
• tsconfig.json is properly configured
• Starter code follows SDK patterns
• .env.example exists with placeholders
• .gitignore protects secrets

Report any issues found.")

Run type checking: !{bash cd $ARGUMENTS && npx tsc --noEmit}

If Python:

Task(description="Verify Python setup", subagent_type="claude-agent-sdk:claude-agent-verifier-py", prompt="Verify the Python Claude Agent SDK setup at $ARGUMENTS.

Check:

• requirements.txt has correct SDK package (claude-agent-sdk)
• Starter code follows SDK patterns
• .env.example exists with placeholders
• .gitignore protects secrets
• Virtual environment is set up

Report any issues found.")

Address any issues found before proceeding

Phase 6: Summary Goal: Provide next steps to user

Actions:

• Summarize what was created:
  • Project structure
  • SDK version installed
  • Files generated
• Provide instructions:
  • How to set API key in .env
  • How to run the agent
  • Links to SDK documentation
• Suggest next steps:
  • Customize system prompt
  • Add custom tools via MCP
  • Create subagents
• Point to examples:
  • Basic usage: examples/python/basic-query.py
  • FastMCP Cloud: examples/python/fastmcp-cloud-http.py
• Common pitfalls to avoid:
  • ✅ Use claude-agent-sdk NOT anthropic-agent-sdk
  • ✅ Use "type": "http" for FastMCP Cloud, NOT "sse"
  • ✅ Pass API keys via env parameter in ClaudeAgentOptions