Plugin Reference

Claude Code plugin API reference. This article provides complete code examples and terminal operation demonstrations.

Quick Start

Start Claude Code in the terminal:

bash

claude

You: Help me implement the Claude Code plugin API reference

Claude: Sure, let me check the project structure first...

Claude wants to:
  - Read file: src/main.py
Allow? [y/n/always/never]

You: y

Claude: Let me implement this feature for you...

Press Shift+Tab to switch permission modes:

bash

# Normal Mode -> Auto-Accept Mode -> Plan Mode

# Auto-Accept: Auto execute (bottom bar shows accept edits on)
You: Quick implementation
Claude: [Auto executes, no confirmation needed]

# Plan Mode: Read-only (bottom bar shows plan mode on)
You: Make a plan first
Claude: [Only makes a plan, does not execute]

Core Concepts

bash

You: Explain the core concepts of the plugin reference

Claude: The core concepts of the Claude Code plugin API reference include...

python

from anthropic import Anthropic

client = Anthropic()
response = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Claude Code plugin API reference"}]
)
print(response.content[0].text)

Implementation Steps

1. Environment Setup

bash

# Install dependencies
pip install anthropic

# Configure API Key
export ANTHROPIC_API_KEY="your-api-key"

2. Core Implementation

bash

You: Help me implement the core plugin reference functionality

Claude: Sure, let me create the implementation file...

Claude wants to:
  - Write file: src/plugin_reference.py
Allow? [y/n/always/never]

You: y

Claude: File created.

python

# Plugin Reference core implementation
from anthropic import Anthropic

client = Anthropic()

def process(query: str) -> str:
    """Process user request"""
    response = client.messages.create(
        model="claude-sonnet-4-6",
        max_tokens=2048,
        messages=[{"role": "user", "content": query}]
    )
    return response.content[0].text

if __name__ == "__main__":
    result = process("Claude Code plugin API reference")
    print(result)

3. Test Verification

bash

You: Run tests to verify the implementation

Claude: Sure, let me run the tests...

Claude wants to:
  - Run command: python -m pytest tests/
Allow? [y/n/always/never]

You: y

Claude: All tests passed

Advanced Usage

python

# Using streaming output
with client.messages.stream(
    model="claude-sonnet-4-6",
    max_tokens=2048,
    messages=[{"role": "user", "content": "Claude Code plugin API reference"}]
) as stream:
    for text in stream.text_stream:
        print(text, end="", flush=True)

bash

# Using tool calls
tools = [{
    "name": "process",
    "description": "Claude Code plugin API reference",
    "input_schema": {
        "type": "object",
        "properties": {
            "query": {"type": "string"}
        },
        "required": ["query"]
    }
}]

response = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    tools=tools,
    messages=[{"role": "user", "content": "Claude Code plugin API reference"}]
)

Performance Optimization

python

# Use Prompt Caching to reduce costs
response = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    system=[{
        "type": "text",
        "text": "You are a Claude Code plugin API reference expert...",
        "cache_control": {"type": "ephemeral"}
    }],
    messages=[...]
)

bash

# Check costs
/cost

# Switch to a faster model
/model haiku

# Use fast mode
/fast

Error Handling

python

from anthropic import APIError, APITimeoutError

try:
    response = client.messages.create(
        model="claude-sonnet-4-6",
        max_tokens=1024,
        messages=[{"role": "user", "content": "request"}]
    )
except APITimeoutError:
    print("Request timed out, please retry")
except APIError as e:
    print(f"API error: {e.status_code} - {e.message}")

Best Practices

Clear task descriptions - Tell Claude specifically what to do
Provide context - Use CLAUDE.md or provide related code directly
Verify step by step - Verify results after each step
Use Plan Mode - Plan first for complex tasks
Monitor costs - Use /cost to check API usage
Choose models wisely - Use Haiku for simple tasks, Opus for complex tasks

Permission Mode Reference

Mode	Shortcut	Bottom Bar	Behavior
Normal	Default	No indicator	Confirm each operation
Auto-Accept	Shift+Tab x1	accept edits on	Auto execute edits
Plan	Shift+Tab x2	plan mode on	Only make plans

Common Commands Reference

bash

/help          # View help
/clear         # Clear history
/model opus    # Switch to Opus 4.6
/model sonnet  # Switch to Sonnet 4.6
/model haiku   # Switch to Haiku 4.5
/cost          # Check costs
/fast          # Fast mode
/quit          # Exit

Plugin Reference ​

Quick Start ​

Core Concepts ​

Implementation Steps ​

1. Environment Setup ​

2. Core Implementation ​

3. Test Verification ​

Advanced Usage ​

Performance Optimization ​

Error Handling ​

Best Practices ​

Permission Mode Reference ​

Common Commands Reference ​

Related Resources ​

Plugin Reference

Quick Start

Core Concepts

Implementation Steps

1. Environment Setup

2. Core Implementation

3. Test Verification

Advanced Usage

Performance Optimization

Error Handling

Best Practices

Permission Mode Reference

Common Commands Reference

Related Resources