Oblien.SearchAgent 1.0.0

Oblien.SearchAgent - Official .NET Client

A clean, production-ready .NET client library for the Oblien Search API. Provides AI-powered web search, intelligent content extraction, and deep website research with real-time streaming capabilities.

Features

• AI-Powered Search - Get relevant web results with optional AI-generated answers
• Smart Extraction - Extract specific content from web pages using natural language
• Deep Research - Crawl and analyze websites with real-time Server-Sent Events (SSE) streaming
• Batch Processing - Process multiple queries or pages simultaneously
• Flexible Authentication - Support for both Client ID/Secret and Bearer token (JWT)
• Clean Architecture - Well-structured, production-ready code with proper error handling
• High Performance - Async/await patterns throughout for optimal performance

Installation

Install via NuGet Package Manager:

dotnet add package Oblien.SearchAgent

Or via Package Manager Console:

Install-Package Oblien.SearchAgent

Or add directly to your .csproj:

<PackageReference Include="Oblien.SearchAgent" Version="1.0.0" />

Quick Start

Using Client ID and Secret

using SearchAgent;
using SearchAgent.Models;

var client = new SearchClient(new SearchClientConfig
{
    ClientId = "your_client_id",
    ClientSecret = "your_client_secret"
});

var results = await client.SearchAsync(
    new[] { "What is TypeScript?" },
    new SearchOptions
    {
        MaxResults = 5,
        IncludeAnswers = true,
        SummaryLevel = "intelligent"
    }
);

foreach (var result in results)
{
    Console.WriteLine($"Query: {result.Query}");
    Console.WriteLine($"Answer: {result.Answer}");
    foreach (var item in result.Results)
    {
        Console.WriteLine($"  - {item.Title}");
        Console.WriteLine($"    {item.Url}");
    }
}

Using Bearer Token

using SearchAgent;
using SearchAgent.Models;

var client = new SearchClient(new SearchClientConfig
{
    BearerToken = "your_jwt_token"
});

var results = await client.SearchAsync(
    new[] { "What is TypeScript?" },
    new SearchOptions
    {
        MaxResults = 5,
        IncludeAnswers = true,
        SummaryLevel = "intelligent"
    }
);

API Methods

Search

Task<List<SearchResult>> SearchAsync(
    string[] queries,
    SearchOptions? options = null,
    CancellationToken cancellationToken = default
)

Search the web with AI-powered answers.

Options:

• IncludeAnswers - Generate AI answers
• SummaryLevel - Quality: 'low', 'medium', 'intelligent'
• MaxResults - Maximum results per query
• Language - Language code (e.g., 'en', 'es')
• Region - Region code (e.g., 'us', 'uk')
• Freshness - Result freshness: 'day', 'week', 'month', 'year'
• IncludeDomains / ExcludeDomains - Domain filtering
• SearchTopic - Topic: 'general', 'news', 'finance'
• SearchDepth - Depth: 'basic', 'advanced'

Extract

Task<ExtractResult> ExtractAsync(
    ExtractPage[] pages,
    ExtractOptions? options = null,
    CancellationToken cancellationToken = default
)

Extract specific content from web pages using natural language instructions.

Example:

var extractResult = await client.ExtractAsync(
    new[]
    {
        new ExtractPage
        {
            Url = "https://example.com/article",
            Details = new[] { "Extract the title", "Extract the author" },
            SummaryLevel = "medium"
        }
    },
    new ExtractOptions
    {
        Format = "markdown",
        Timeout = 30000
    }
);

Options:

• Format - Output format: 'markdown', 'html', 'text'
• ExtractDepth - Depth: 'basic', 'advanced'
• IncludeImages / IncludeFavicon - Include images/favicon
• MaxContentLength - Maximum content length
• Timeout - Request timeout in milliseconds

Crawl

Task<CrawlResult> CrawlAsync(
    string instructions,
    CrawlOptions? options = null,
    Action<CrawlEvent>? onEvent = null,
    CancellationToken cancellationToken = default
)

Crawl and research websites with real-time streaming updates.

Example:

var crawlResult = await client.CrawlAsync(
    "Research https://example.com and summarize key features",
    new CrawlOptions
    {
        Type = "deep",
        MaxPages = 10,
        Thinking = true
    },
    onEvent: (evt) =>
    {
        Console.WriteLine($"[{evt.Type}] {evt.Data}");
    }
);

Options:

• Type - Crawl type: 'deep', 'shallow', 'focused'
• MaxDepth / MaxPages - Crawl limits
• IncludeExternal - Include external links
• Thinking - Enable AI thinking process
• StreamText - Stream text results
• Format - Output format

Architecture

SearchAgent/
├── Models/
│   ├── SearchOptions.cs
│   ├── SearchResult.cs
│   ├── ExtractOptions.cs
│   ├── ExtractPage.cs
│   ├── ExtractResult.cs
│   ├── CrawlOptions.cs
│   ├── CrawlResult.cs
│   └── CrawlEvent.cs
├── SearchClient.cs
└── SearchClientConfig.cs

Testing

Run the test application with Client ID/Secret:

cd tests/SearchAgent.Tests
dotnet run

Test with Bearer Token:

cd tests/SearchAgent.Tests
dotnet run -- --bearer "your_jwt_token"

Authentication

The client supports two authentication methods:

1. Client ID and Secret (Header-based)

Best for server-to-server communication:

var client = new SearchClient(new SearchClientConfig
{
    ClientId = "oblien_...",
    ClientSecret = "sk_..."
});

Get your credentials from the Oblien Dashboard.

2. Bearer Token (JWT-based)

Required for Crawl API with SSE streaming:

var client = new SearchClient(new SearchClientConfig
{
    BearerToken = "eyJhbGc..."
});

Important: The Crawl API endpoint requires Bearer token authentication for real-time SSE streaming to work properly.

Requirements

• .NET 8.0 or higher
• Oblien API credentials (Client ID and Secret)

License

This is a client library for the Oblien Search API. Refer to Oblien's terms of service for API usage.