Commands & Snippets

Fresh

Quick reference for common commands and code patterns.

Claude API

Installation

# Python
pip install anthropic

# Node.js
npm install @anthropic-ai/sdk

Environment Setup

# Set API key
export ANTHROPIC_API_KEY="sk-ant-api03-..."

# Windows PowerShell
$env:ANTHROPIC_API_KEY = "sk-ant-api03-..."

Basic Messages

import anthropic

client = anthropic.Anthropic()

# Simple message
response = client.messages.create(
    model="claude-sonnet-4-20250514",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello!"}]
)

# With system prompt
response = client.messages.create(
    model="claude-sonnet-4-20250514",
    max_tokens=1024,
    system="You are a helpful assistant.",
    messages=[{"role": "user", "content": "Hello!"}]
)

Streaming

with client.messages.stream(
    model="claude-sonnet-4-20250514",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Tell me a story"}]
) as stream:
    for text in stream.text_stream:
        print(text, end="", flush=True)

Tool Use

tools = [{
    "name": "get_weather",
    "description": "Get weather for a location",
    "input_schema": {
        "type": "object",
        "properties": {
            "location": {"type": "string"}
        },
        "required": ["location"]
    }
}]

response = client.messages.create(
    model="claude-sonnet-4-20250514",
    max_tokens=1024,
    tools=tools,
    messages=[{"role": "user", "content": "Weather in Boston?"}]
)

# Check for tool use
for block in response.content:
    if block.type == "tool_use":
        print(f"Tool: {block.name}")
        print(f"Input: {block.input}")

Claude Code CLI

Session Commands

# Start new session
claude

# Resume last session
claude --resume

# Continue with context
claude --continue

# Start with prompt
claude "explain this code"

In-Session Commands

/help         - Show help
/clear        - Clear context
/exit         - Exit session
/compact      - Compress context

Common Prompts

"Look at [file] and explain how it works"
"Create a new [component/function] that does [X]"
"Fix the error in [file]"
"Write tests for [module]"
"Refactor this to use [pattern]"
"Create a commit for these changes"

MCP Commands

Server Development

# Install MCP SDK
pip install mcp

# Run server directly
python server.py

# Run with uvicorn (HTTP)
uvicorn server:app --reload

Client Configuration

{
  "mcpServers": {
    "my-server": {
      "command": "python",
      "args": ["path/to/server.py"],
      "env": {
        "API_KEY": "your-key"
      }
    }
  }
}

Config File Locations

Platform	Location
macOS	~/Library/Application Support/Claude/claude_desktop_config.json
Windows	%APPDATA%\Claude\claude_desktop_config.json
Linux	~/.config/Claude/claude_desktop_config.json

RAG Patterns

Chunking

def chunk_text(text, size=500, overlap=50):
    chunks = []
    start = 0
    while start < len(text):
        end = start + size
        chunks.append(text[start:end])
        start = end - overlap
    return chunks

Embedding

import voyageai

client = voyageai.Client()

# Embed documents
embeddings = client.embed(
    texts=chunks,
    model="voyage-3",
    input_type="document"
).embeddings

# Embed query
query_embedding = client.embed(
    texts=[query],
    model="voyage-3",
    input_type="query"
).embeddings[0]

Hybrid Search

from rank_bm25 import BM25Okapi
from sklearn.metrics.pairwise import cosine_similarity

def hybrid_search(query, chunks, embeddings):
    # Semantic
    q_emb = embed_query(query)
    semantic_scores = cosine_similarity([q_emb], embeddings)[0]

    # BM25
    bm25 = BM25Okapi([c.split() for c in chunks])
    bm25_scores = bm25.get_scores(query.split())

    # Combine (RRF)
    return reciprocal_rank_fusion(semantic_scores, bm25_scores)

Model Reference

Model	Best For
claude-opus-4-5-20251101	Complex reasoning, analysis
claude-sonnet-4-20250514	Balanced performance
claude-haiku-3-20240307	Fast, simple tasks

Token Limits

Model	Context	Max Output
Claude 3.5 Sonnet	200K	8K
Claude 3 Opus	200K	4K
Claude 3 Haiku	200K	4K