+++++

Snippets(6)

tsQuick Start

Minimal setup — import the SDK and send your first message.

import Anthropic from '@anthropic-ai/sdk'

const client = new Anthropic()

const msg = await client.messages.create({
  model: 'claude-opus-4-8',
  max_tokens: 1024,
  messages: [{ role: 'user', content: 'Hello!' }],
})

console.log(msg.content[0].text)

tsMulti-turn Conversation

Accumulate messages across turns to build a stateful conversation.

import { MessageParam } from '@anthropic-ai/sdk/resources'

const history: MessageParam[] = []

history.push({ role: 'user', content: 'My name is David.' })
const r1 = await client.messages.create({
  model: 'claude-opus-4-8',
  max_tokens: 256,
  messages: history,
})
history.push({ role: 'assistant', content: r1.content })

history.push({ role: 'user', content: 'What is my name?' })
const r2 = await client.messages.create({
  model: 'claude-opus-4-8',
  max_tokens: 256,
  messages: history,
})
console.log(r2.content[0].text) // "Your name is David."

tsSystem Prompt

Shape Claude's persona and constraints before the conversation starts.

const msg = await client.messages.create({
  model: 'claude-opus-4-8',
  max_tokens: 1024,
  system: `You are a concise coding assistant.
Be direct. Prefer TypeScript. No filler.`,
  messages: [{
    role: 'user',
    content: 'Write a debounce function.',
  }],
})

console.log(msg.content[0].text)

tsPrompt Caching

Cache a large system prompt to cut latency and cost on repeated calls.

const msg = await client.messages.create({
  model: 'claude-opus-4-8',
  max_tokens: 1024,
  system: [{
    type: 'text',
    text: systemPrompt, // your large prompt
    cache_control: { type: 'ephemeral' },
  }],
  messages: [{ role: 'user', content: 'Summarize the key points.' }],
})

const { cache_creation_input_tokens, cache_read_input_tokens } = msg.usage
console.log(`Cache write: ${cache_creation_input_tokens ?? 0}`)
console.log(`Cache read:  ${cache_read_input_tokens ?? 0}`)

tsTool Use

Give Claude a tool it can call — here, a weather lookup.

const msg = await client.messages.create({
  model: 'claude-opus-4-8',
  max_tokens: 1024,
  tools: [{
    name: 'get_weather',
    description: 'Get weather for a location',
    input_schema: {
      type: 'object',
      properties: {
        location: { type: 'string' },
      },
      required: ['location'],
    },
  }],
  messages: [{
    role: 'user',
    content: 'What is the weather in Tokyo?',
  }],
})

tsStreaming

Stream tokens as they arrive instead of waiting for the full reply.

const stream = client.messages.stream({
  model: 'claude-opus-4-8',
  max_tokens: 1024,
  messages: [{
    role: 'user',
    content: 'Explain closures in one paragraph.',
  }],
})

for await (const chunk of stream) {
  if (chunk.type === 'content_block_delta') {
    process.stdout.write(chunk.delta.text)
  }
}

const final = await stream.finalMessage()

/ BLOGWriting on AI, engineering, and building with ClaudeVIEW ALL →