### Install Frontend Dependencies

Source: https://github.com/aws-samples/amazon-bedrock-rag/blob/main/README.md

Run this command in the frontend folder to install necessary packages for the frontend application.

```bash
npm install
```

--------------------------------

### Start Development Server with npm start

Source: https://github.com/aws-samples/amazon-bedrock-rag/blob/main/frontend/README.md

Runs the app in development mode. The server automatically reloads on changes and displays lint errors in the console. Accessible at http://localhost:3000.

```bash
npm start
```

--------------------------------

### Launch Frontend Application

Source: https://github.com/aws-samples/amazon-bedrock-rag/blob/main/README.md

Execute this command to start the frontend application and access it via your browser.

```bash
npm run start
```

--------------------------------

### Deploy Backend Stack with CDK

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Install dependencies, deploy the CDK stack with an IP whitelist, and upload documents to S3. This triggers automatic ingestion into the knowledge base.

```bash
# Install dependencies
cd backend
npm install
```

```bash
# Deploy with IP whitelist (required)
cdk deploy --context allowedip="203.0.113.0/32"
```

```bash
# Upload documents to S3 (triggers automatic ingestion)
aws s3 cp Amazon-2022-Annual-Report.pdf s3://docsbucket-uuid-12345/
```

```bash
# Manually sync web crawler (first time)
aws bedrock-agent start-ingestion-job \
  --knowledge-base-id KBXXXXXXXX \
  --data-source-id DSXXXXXXXX
```

--------------------------------

### Query Knowledge Base with Session Support

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Handles question-answering using Amazon Bedrock's RetrieveAndGenerate API with knowledge base integration and session management for multi-turn conversations. Pass null for sessionId to start a new session.

```javascript
const {
  BedrockAgentRuntimeClient,
  RetrieveAndGenerateCommand,
} = require("@aws-sdk/client-bedrock-agent-runtime");

const client = new BedrockAgentRuntimeClient({ region: "us-east-1" });

// Query with session support for multi-turn conversations
const queryKnowledgeBase = async (question, modelId, sessionId) => {
  const input = {
    sessionId: sessionId, // Pass existing sessionId for conversation context
    input: {
      text: question,
    },
    retrieveAndGenerateConfiguration: {
      type: "KNOWLEDGE_BASE",
      knowledgeBaseConfiguration: {
        knowledgeBaseId: process.env.KNOWLEDGE_BASE_ID,
        modelArn: `arn:aws:bedrock:us-east-1::foundation-model/${modelId}`,
      },
    },
  };

  const command = new RetrieveAndGenerateCommand(input);
  const response = await client.send(command);

  // Extract citation based on source type (S3 or Web)
  const location = response.citations[0]?.retrievedReferences[0]?.location;
  const citation = location?.type === "S3"
    ? location.s3Location.uri
    : location?.webLocation?.url;

  return {
    response: response.output.text,
    citation: citation,
    sessionId: response.sessionId,
  };
};

// Usage example
const result = await queryKnowledgeBase(
  "What is Amazon's revenue?",
  "anthropic.claude-3-haiku-20240307-v1:0",
  null // null for new session
);
console.log(result);
// { response: "Amazon's revenue was...", citation: "s3://bucket/report.pdf", sessionId: "new-session-id" }
```

--------------------------------

### Get Web Crawler URLs - GET /urls

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Retrieves the current configuration of the web crawler data source, including seed URLs, exclusion filters, and inclusion filters.

```bash
curl -X GET "https://<api-gateway-url>/urls" \
  -H "Content-Type: application/json"
```

```json
{
  "seedUrlList": [
    { "url": "https://www.aboutamazon.com/news/amazon-offices" },
    { "url": "https://www.aboutamazon.com/news/amazon-offices/amazon-headquarters-tour-seattle" }
  ],
  "exclusionFilters": [".*plants.*"],
  "inclusionFilters": ["^https?://www.aboutamazon.com/news/amazon-offices/.*$"]
}
```

--------------------------------

### Start Knowledge Base Ingestion Job

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Triggers knowledge base ingestion for S3 data sources. This function is intended to be automatically invoked via S3 event triggers.

```javascript
const {
  BedrockAgentClient,
  StartIngestionJobCommand,
} = require("@aws-sdk/client-bedrock-agent");

const client = new BedrockAgentClient({ region: "us-east-1" });

// Start ingestion job for S3 data source
const startIngestion = async (knowledgeBaseId, dataSourceId, clientToken) => {
  const input = {
    knowledgeBaseId: knowledgeBaseId,
    dataSourceId: dataSourceId,
    clientToken: clientToken, // Idempotency token from Lambda context
  };

  const command = new StartIngestionJobCommand(input);
  const response = await client.send(command);

  return {
    ingestionJob: response.ingestionJob,
  };
};

// Triggered automatically on S3 PUT events
exports.handler = async (event, context) => {
  return await startIngestion(
    process.env.KNOWLEDGE_BASE_ID,
    process.env.DATA_SOURCE_ID,
    context.awsRequestId
  );
};
```

--------------------------------

### Get Web Crawler URLs

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Retrieves the current configuration of the web crawler data source including seed URLs, exclusion filters, and inclusion filters.

```APIDOC
## GET /urls

### Description
Retrieves the current configuration of the web crawler data source including seed URLs, exclusion filters, and inclusion filters.

### Method
GET

### Endpoint
/urls

### Response
#### Success Response (200)
- **seedUrlList** (array) - A list of seed URLs for the web crawler.
  - **url** (string) - The URL.
- **exclusionFilters** (array) - A list of regular expressions for excluding URLs.
- **inclusionFilters** (array) - A list of regular expressions for including URLs.

#### Response Example
```json
{
  "seedUrlList": [
    { "url": "https://www.aboutamazon.com/news/amazon-offices" },
    { "url": "https://www.aboutamazon.com/news/amazon-offices/amazon-headquarters-tour-seattle" }
  ],
  "exclusionFilters": [".*plants.*"],
  "inclusionFilters": ["^https?://www.aboutamazon.com/news/amazon-offices/.*$"]
}
```
```

--------------------------------

### Build for Production with npm run build

Source: https://github.com/aws-samples/amazon-bedrock-rag/blob/main/frontend/README.md

Builds the application for production, optimizing it for performance. The output is minified and placed in the 'build' folder, ready for deployment.

```bash
npm run build
```

--------------------------------

### Check Frontend Vulnerabilities

Source: https://github.com/aws-samples/amazon-bedrock-rag/blob/main/README.md

Optionally, run this command to check for development vulnerabilities in the frontend dependencies.

```bash
npm audit --omit=dev
```

--------------------------------

### Deploy Backend with AWS CDK

Source: https://github.com/aws-samples/amazon-bedrock-rag/blob/main/README.md

Use AWS CDK to deploy the backend infrastructure. Provide an allowed client IP address for API Gateway access using the 'allowedip' context variable.

```bash
cdk deploy --context allowedip="xxx.xxx.xxx.xxx/32"
```

--------------------------------

### Run Tests with npm test

Source: https://github.com/aws-samples/amazon-bedrock-rag/blob/main/frontend/README.md

Launches the test runner in interactive watch mode. Refer to the Create React App documentation for more information on running tests.

```bash
npm test
```

--------------------------------

### Query Knowledge Base - POST /docs

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Sends a question to the RAG system to retrieve an AI-generated answer with citations. Supports multi-turn conversations using session IDs and model selection.

```bash
curl -X POST "https://<api-gateway-url>/docs" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "Which cities have Amazon offices?",
    "modelId": "anthropic.claude-3-haiku-20240307-v1:0",
    "requestSessionId": null
  }'
```

```json
{
  "response": "Amazon has offices in several cities including Seattle, New York, and San Francisco...",
  "citation": "s3://docs-bucket/Amazon-2022-Annual-Report.pdf",
  "sessionId": "abc123-session-id"
}
```

```bash
curl -X POST "https://<api-gateway-url>/docs" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "What about European offices?",
    "modelId": "anthropic.claude-3-haiku-20240307-v1:0",
    "requestSessionId": "abc123-session-id"
  }'
```

```json
{
  "response": "Amazon has offices in London, Luxembourg, and Dublin...",
  "citation": "https://www.aboutamazon.com/news/amazon-offices/european-headquarters",
  "sessionId": "abc123-session-id"
}
```

--------------------------------

### Create Web Crawler Data Source

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Creates a new web crawler data source for indexing website content into a knowledge base. Configures inclusion/exclusion filters and crawler limits.

```javascript
const {
  BedrockAgentClient,
  CreateDataSourceCommand,
} = require("@aws-sdk/client-bedrock-agent");

const client = new BedrockAgentClient({ region: "us-east-1" });

// Create a new web crawler data source
const createWebDataSource = async (knowledgeBaseId) => {
  const input = {
    knowledgeBaseId: knowledgeBaseId,
    name: "WebCrawlerDataSource",
    dataSourceConfiguration: {
      type: "WEB",
      webConfiguration: {
        sourceConfiguration: {
          urlConfiguration: {
            seedUrls: [
              { url: "https://www.aboutamazon.com/news/amazon-offices" },
              { url: "https://docs.example.com/" },
            ],
          },
        },
        crawlerConfiguration: {
          crawlerLimits: { rateLimit: 50 },
          scope: "HOST_ONLY",
          exclusionFilters: [".*plants.*", ".*archive.*"],
          inclusionFilters: ["^https?://www.aboutamazon.com/news/.*$"],
        },
      },
    },
    dataDeletionPolicy: "DELETE",
  };

  const command = new CreateDataSourceCommand(input);
  const response = await client.send(command);
  return response.dataSource;
};
```

--------------------------------

### Query Knowledge Base

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Sends a question to the RAG system and retrieves an AI-generated answer with citations from the knowledge base. Supports multi-turn conversations via session management and allows selecting different foundation models.

```APIDOC
## POST /docs

### Description
Sends a question to the RAG system and retrieves an AI-generated answer with citations from the knowledge base. The endpoint supports multi-turn conversations via session management and allows selecting different foundation models.

### Method
POST

### Endpoint
/docs

### Request Body
- **question** (string) - Required - The user's question.
- **modelId** (string) - Required - The ID of the foundation model to use for generation.
- **requestSessionId** (string) - Optional - The session ID for continuing a conversation.

### Request Example
```json
{
  "question": "Which cities have Amazon offices?",
  "modelId": "anthropic.claude-3-haiku-20240307-v1:0",
  "requestSessionId": null
}
```

### Response
#### Success Response (200)
- **response** (string) - The AI-generated answer.
- **citation** (string) - A citation for the source of the answer.
- **sessionId** (string) - The session ID for the current conversation.

#### Response Example
```json
{
  "response": "Amazon has offices in several cities including Seattle, New York, and San Francisco...",
  "citation": "s3://docs-bucket/Amazon-2022-Annual-Report.pdf",
  "sessionId": "abc123-session-id"
}
```
```

--------------------------------

### React Frontend Chat Interface Integration

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Integrates a React frontend with the backend API Gateway for a chat interface. Handles conversation history, session management, model selection, and document/web URL updates. Ensure the `baseUrl` is correctly configured for your deployed API.

```javascript
import { useState } from "react";

const App = () => {
  const [history, setHistory] = useState([]);
  const [sessionId, setSessionId] = useState(undefined);
  const [selectedModel, setSelectedModel] = useState({
    modelId: "anthropic.claude-3-haiku-20240307-v1:0",
    modelName: "Claude 3 Haiku",
  });

  // Available models for selection
  const modelList = [
    { modelId: "anthropic.claude-3-haiku-20240307-v1:0", modelName: "Claude 3 Haiku" },
    { modelId: "meta.llama3-70b-instruct-v1:0", modelName: "Llama 3 70B Instruct" },
  ];

  const handleSendQuestion = async (question, baseUrl) => {
    const response = await fetch(`${baseUrl}docs`, {
      method: "POST",
      headers: { "Content-Type": "application/json" },
      body: JSON.stringify({
        requestSessionId: sessionId,
        question: question,
        modelId: selectedModel.modelId,
      }),
    });

    const data = await response.json();
    setSessionId(data.sessionId); // Maintain session for multi-turn

    setHistory([
      ...history,
      {
        question: question,
        response: data.response,
        citation: data.citation,
      },
    ]);
  };

  // Fetch and update web crawler configuration
  const handleUpdateUrls = async (urls, exclusionFilters, inclusionFilters, baseUrl) => {
    const response = await fetch(`${baseUrl}web-urls`, {
      method: "POST",
      headers: { "Content-Type": "application/json" },
      body: JSON.stringify({
        urlList: [...new Set(urls)], // Deduplicate
        exclusionFilters: [...new Set(exclusionFilters)],
        inclusionFilters: [...new Set(inclusionFilters)],
      }),
    });
    return response.ok;
  };

  return (/* JSX */);
};
```

--------------------------------

### Configure Knowledge Base with CDK

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Defines a vector knowledge base using Amazon Titan Embeddings and configures an S3 data source with a fixed-size chunking strategy. Ensure `RemovalPolicy.DESTROY` and `autoDeleteObjects: true` are used cautiously in production.

```typescript
import { bedrock } from "@cdklabs/generative-ai-cdk-constructs";
import * as s3 from "aws-cdk-lib/aws-s3";

// Create Knowledge Base with vector embeddings
const knowledgeBase = new bedrock.VectorKnowledgeBase(this, "knowledgeBase", {
  embeddingsModel: bedrock.BedrockFoundationModel.TITAN_EMBED_TEXT_V2_1024,
  vectorType: bedrock.VectorType.BINARY,
});

// Create S3 bucket for documents
const docsBucket = new s3.Bucket(this, "docsbucket", {
  blockPublicAccess: s3.BlockPublicAccess.BLOCK_ALL,
  encryption: s3.BucketEncryption.S3_MANAGED,
  enforceSSL: true,
  removalPolicy: RemovalPolicy.DESTROY,
  autoDeleteObjects: true,
  lifecycleRules: [{ expiration: Duration.days(10) }],
});

// Configure S3 data source with chunking strategy
const s3DataSource = new bedrock.S3DataSource(this, "s3DataSource", {
  bucket: docsBucket,
  knowledgeBase: knowledgeBase,
  dataSourceName: "docs",
  chunkingStrategy: bedrock.ChunkingStrategy.fixedSize({
    maxTokens: 500,
    overlapPercentage: 20,
  }),
});
```

--------------------------------

### Eject Configuration with npm run eject

Source: https://github.com/aws-samples/amazon-bedrock-rag/blob/main/frontend/README.md

Permanently removes the single build dependency from your project, copying all configuration files (webpack, Babel, ESLint) into your project for full control. This is a one-way operation.

```bash
npm run eject
```

--------------------------------

### Destroy Cloud Resources

Source: https://github.com/aws-samples/amazon-bedrock-rag/blob/main/README.md

Use this command to remove the AWS cloud resources created during the solution deployment.

```bash
cdk destroy
```

--------------------------------

### Update Web Crawler URLs - POST /web-urls

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Updates the web crawler data source configuration with new seed URLs and filter patterns. Changes are applied during the next scheduled sync.

```bash
curl -X POST "https://<api-gateway-url>/web-urls" \
  -H "Content-Type: application/json" \
  -d '{
    "urlList": [
      "https://www.aboutamazon.com/news/amazon-offices",
      "https://www.aboutamazon.com/news/amazon-offices/amazon-headquarters-tour-seattle",
      "https://docs.aws.amazon.com/bedrock/latest/userguide/"
    ],
    "exclusionFilters": [".*plants.*", ".*deprecated.*"],
    "inclusionFilters": ["^https?://www.aboutamazon.com/news/amazon-offices/.*$"]
  }'
```

```json
{
  "dataSource": {
    "dataSourceId": "DATASOURCE123",
    "knowledgeBaseId": "KBXXXXXXXX",
    "name": "WebCrawlerDataSource",
    "status": "AVAILABLE"
  }
}
```

--------------------------------

### Update Web Crawler Data Source Configuration

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Updates an existing web crawler data source, allowing modification of seed URLs, inclusion filters, and exclusion filters. Ensures idempotency with a client token.

```javascript
const {
  BedrockAgentClient,
  UpdateDataSourceCommand,
} = require("@aws-sdk/client-bedrock-agent");

const client = new BedrockAgentClient({ region: "us-east-1" });

// Update seed URLs and filters
const updateWebDataSource = async (knowledgeBaseId, dataSourceId, urlList, exclusionFilters, inclusionFilters) => {
  const seedUrls = urlList.map((url) => ({ url: url }));

  const input = {
    knowledgeBaseId: knowledgeBaseId,
    dataSourceId: dataSourceId,
    name: "WebCrawlerDataSource",
    dataSourceConfiguration: {
      type: "WEB",
      webConfiguration: {
        sourceConfiguration: {
          urlConfiguration: { seedUrls: seedUrls },
        },
        crawlerConfiguration: {
          crawlerLimits: { rateLimit: 50 },
          scope: "HOST_ONLY",
          exclusionFilters: exclusionFilters,
          inclusionFilters: inclusionFilters,
        },
      },
    },
    dataDeletionPolicy: "DELETE",
  };

  const command = new UpdateDataSourceCommand(input);
  return await client.send(command);
};
```

--------------------------------

### Update Web Crawler URLs

Source: https://context7.com/aws-samples/amazon-bedrock-rag/llms.txt

Updates the web crawler data source configuration with new seed URLs and filter patterns. Changes take effect on the next scheduled sync (daily).

```APIDOC
## POST /web-urls

### Description
Updates the web crawler data source configuration with new seed URLs and filter patterns. Changes take effect on the next scheduled sync (daily).

### Method
POST

### Endpoint
/web-urls

### Request Body
- **urlList** (array) - Required - A list of seed URLs for the web crawler.
- **exclusionFilters** (array) - Optional - A list of regular expressions for excluding URLs.
- **inclusionFilters** (array) - Optional - A list of regular expressions for including URLs.

### Request Example
```json
{
  "urlList": [
    "https://www.aboutamazon.com/news/amazon-offices",
    "https://www.aboutamazon.com/news/amazon-offices/amazon-headquarters-tour-seattle",
    "https://docs.aws.amazon.com/bedrock/latest/userguide/"
  ],
  "exclusionFilters": [".*plants.*", ".*deprecated.*"],
  "inclusionFilters": ["^https?://www.aboutamazon.com/news/amazon-offices/.*$"]
}
```

### Response
#### Success Response (200)
- **dataSource** (object) - Information about the updated data source.
  - **dataSourceId** (string) - The ID of the data source.
  - **knowledgeBaseId** (string) - The ID of the knowledge base.
  - **name** (string) - The name of the data source.
  - **status** (string) - The status of the data source.

#### Response Example
```json
{
  "dataSource": {
    "dataSourceId": "DATASOURCE123",
    "knowledgeBaseId": "KBXXXXXXXX",
    "name": "WebCrawlerDataSource",
    "status": "AVAILABLE"
  }
}
```
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.