快速开始

本指南将引导你完成 Hindsight Cloud 账户的设置，并创建你的第一个记忆库。

前置条件

• 一个 Hindsight Cloud 账户（在 ui.hindsight.vectorize.io 注册）
• 对 REST API 有基本了解，或熟悉 Python/TypeScript

第 1 步：注册并登录

1. 访问 ui.hindsight.vectorize.io
2. 点击 Sign Up 创建账户
3. 验证你的邮箱地址
4. 登录以进入仪表盘

第 2 步：创建组织

首次登录时，系统会提示你创建一个组织。组织是所有资源的顶层容器：

• 记忆库
• 团队成员
• API 密钥
• 账单信息

输入组织名称，然后点击 Create。

第 3 步：创建你的第一个记忆库

1. 在仪表盘中点击 Create Memory Bank
2. 为记忆库输入名称（例如 "My AI Assistant"）

第 4 步：获取 API 密钥

1. 点击顶部导航栏的 Connect 按钮
2. 点击 Create API Key
3. 为密钥起一个有描述性的名称
4. 选择过期时间（或选择 "Never"）
5. 重要：立即复制你的 API 密钥——它不会再次显示！

第 5 步：发起你的第一次 API 调用

Python

pip install hindsight-client

from hindsight_client import Hindsight

client = Hindsight(
    base_url="https://api.hindsight.vectorize.io",
    api_key="your-api-key"
)

# Create a memory bank
bank = client.create_bank(
    bank_id="my-assistant",
    name="My Assistant"
)

# Store a memory
client.retain(
    bank_id="my-assistant",
    content="The user's name is Alice and she prefers Python."
)

# Retrieve relevant memories
result = client.recall(
    bank_id="my-assistant",
    query="What is the user's name?"
)

for memory in result.results:
    print(memory.text)

# Create a mental model - a pre-computed reflection that stays current
result = client.create_mental_model(
    bank_id="my-assistant",
    name="User Profile",
    source_query="What do we know about this user?"
)
print(f"Creating mental model (operation: {result.operation_id})")

# List mental models
models = client.list_mental_models(bank_id="my-assistant")
for m in models.items:
    print(f"{m.name}: {m.content}")

client.close()

TypeScript

npm install @vectorize-io/hindsight-client

import { HindsightClient } from '@vectorize-io/hindsight-client';

const client = new HindsightClient({
  baseUrl: 'https://api.hindsight.vectorize.io',
  apiKey: 'your-api-key'
});

// Create a memory bank
const bank = await client.createBank('my-assistant', {
  name: 'My Assistant'
});

// Store a memory
await client.retain(
  'my-assistant',
  "The user's name is Alice and she prefers Python."
);

// Retrieve relevant memories
const result = await client.recall(
  'my-assistant',
  "What is the user's name?"
);

result.results.forEach(memory => {
  console.log(memory.text);
});

// Create a mental model - a pre-computed reflection that stays current
const mmResult = await client.createMentalModel(
  'my-assistant',
  'User Profile',
  'What do we know about this user?'
);
console.log(`Creating mental model (operation: ${mmResult.operation_id})`);

// List mental models
const models = await client.listMentalModels('my-assistant');
models.items.forEach(m => {
  console.log(`${m.name}: ${m.content}`);
});

cURL

# Store a memory
curl -X POST https://api.hindsight.vectorize.io/v1/default/banks/{bank_id}/memories \
  -H "Authorization: Bearer your-api-key" \
  -H "Content-Type: application/json" \
  -d '{"items": [{"content": "The user'\''s name is Alice and she prefers Python."}]}'

# Retrieve memories
curl -X POST https://api.hindsight.vectorize.io/v1/default/banks/{bank_id}/memories/recall \
  -H "Authorization: Bearer your-api-key" \
  -H "Content-Type: application/json" \
  -d '{"query": "What is the user'\''s name?"}'

# Create a mental model - a pre-computed reflection that stays current
curl -X POST https://api.hindsight.vectorize.io/v1/default/banks/{bank_id}/mental-models \
  -H "Authorization: Bearer your-api-key" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "User Profile",
    "source_query": "What do we know about this user?"
  }'

后续步骤

• 深入了解 记忆操作
• 创建 心智模型 以从记忆中整理出预计算的反思
• 设置 团队管理 以便协作
• 配置 账单 并购买额度
• 浏览 API 集成 指南以进行高级用法