Kode Agent SDK for .NET

中文版: 查看中文 README

A powerful AI Agent runtime SDK built on .NET 10, supporting Anthropic Claude and OpenAI GPT models, with comprehensive capabilities for tool calling, state management, event streaming, and more. Designed for building production-grade AI applications.

Project Background

Kode Agent SDK for .NET is the C# implementation of Kode SDK, providing feature parity with the TypeScript version:

Event-Driven Architecture - Three-channel event system (Progress, Control, Monitor) for separation of concerns
Complete State Management - Support for persistence, crash recovery, and resumable execution
Flexible Tool System - 20+ built-in tools with support for custom tools and MCP protocol
Multi-Model Support - Anthropic Claude, OpenAI GPT, and compatible APIs
Enterprise Features - Permission control, sandbox execution, dependency injection, logging integration

Features

Feature	Description	Documentation
Multi-Model Support	Support for Anthropic Claude (Claude 4, 3.5) and OpenAI GPT (GPT-4o, o1) series models, plus compatible APIs	Model Providers
Rich Tool System	20+ built-in tools (file system, shell, todo management, etc.), support for custom tools and MCP protocol	Tool Development Guide
Three-Channel Event Stream	Progress (real-time output), Control (approval flow), Monitor (logging) separated architecture	Event System Details
State Persistence	JSON file storage and Redis distributed storage, support for session recovery and resumable execution	State Storage
Permission Control	Fine-grained tool permission management, support for auto-approval, manual approval, and deny policies	Permission Control System
Sandbox Execution	Secure command execution environment, support for local and Docker sandboxes	Advanced Guide - Sandbox
Source Generator	Compile-time tool schema generation, zero reflection overhead, type-safe	Tool Development Guide
MCP Integration	Native Model Context Protocol support, easy integration with external tool ecosystem	MCP Protocol Integration
Dependency Injection	Complete Microsoft.Extensions.DependencyInjection support	Dependency Injection
Template System	Predefined agent templates for quick creation of scenario-specific agents	Sub-Agent Delegation
Skills System	Progressive skill discovery and activation, dynamic extension of agent capabilities	Skills System
Sub-Agent Delegation	Support for task delegation to specialized sub-agents, enabling complex workflow orchestration	Sub-Agent Delegation

Architecture Overview

graph TB
    subgraph Application["Application Layer"]
        App[Your Application]
    end

    subgraph Agent["Agent Core"]
        AgentCore[Agent]
        Config[AgentConfig]
        State[Runtime State]
        EventBus[EventBus]
        Skills[SkillsManager]
    end

    subgraph Dependencies["Dependencies"]
        Store[(AgentStore)]
        Registry[ToolRegistry]
        Provider[ModelProvider]
        Sandbox[Sandbox]
    end

    subgraph Tools["Tools"]
        Builtin[Builtin Tools]
        Custom[Custom Tools]
        MCP[MCP Tools]
        SkillTools[Skill Tools]
        TaskRun[task_run<br/>Sub-Agent]
    end

    subgraph Providers["Model Providers"]
        Anthropic[Anthropic Claude]
        OpenAI[OpenAI GPT]
        Compatible[Compatible APIs]
    end

    subgraph Storage["Storage"]
        JSON[(JSON Store)]
        Redis[(Redis Store)]
    end

    App --> AgentCore
    AgentCore --> Config
    AgentCore --> State
    AgentCore --> EventBus
    AgentCore --> Skills

    AgentCore --> Store
    AgentCore --> Registry
    AgentCore --> Provider
    AgentCore --> Sandbox

    Registry --> Builtin
    Registry --> Custom
    Registry --> MCP
    Registry --> SkillTools
    Registry --> TaskRun

    Provider --> Anthropic
    Provider --> OpenAI
    Provider --> Compatible

    Store --> JSON
    Store --> Redis

Agent Execution Flow

sequenceDiagram
    participant User
    participant Agent
    participant EventBus
    participant Provider as Model Provider
    participant Tools as Tool Registry

    User->>Agent: RunAsync("User message")
    Agent->>EventBus: Publish(Progress, Start)

    loop Agent Loop
        Agent->>Provider: StreamAsync(messages)
        Provider-->>Agent: Text chunks
        Agent->>EventBus: Publish(Progress, TextChunk)

        alt Tool Call Required
            Provider-->>Agent: ToolUse
            Agent->>EventBus: Publish(Progress, ToolStart)

            alt Requires Approval
                Agent->>EventBus: Publish(Control, ApprovalRequest)
                EventBus-->>Agent: ApprovalResponse
            end

            Agent->>Tools: ExecuteAsync(tool, args)
            Tools-->>Agent: ToolResult
            Agent->>EventBus: Publish(Progress, ToolEnd)
        end
    end

    Agent->>EventBus: Publish(Progress, Done)
    Agent-->>User: AgentRunResult

Quick Start

1. Install Packages

From NuGet.org (Recommended):

# Core packages
dotnet add package Kode.Agent.Sdk
dotnet add package Kode.Agent.Store.Json
dotnet add package Kode.Agent.Tools.Builtin

# Optional: MCP support
dotnet add package Kode.Agent.Mcp

# Optional: Redis storage
dotnet add package Kode.Agent.Store.Redis

From GitHub Packages:

# Configure GitHub Packages source (one-time setup)
dotnet nuget add source \
  --username YOUR_GITHUB_USERNAME \
  --password YOUR_GITHUB_TOKEN \
  --store-password-in-clear-text \
  --name github \
  "https://nuget.pkg.github.com/JinFanZheng/index.json"

# Install packages
dotnet add package Kode.Agent.Sdk --source github

📚 For detailed GitHub Packages setup, see GitHub Packages Guide

Or reference projects directly:

<ProjectReference Include="path/to/Kode.Agent.Sdk.csproj" />
<ProjectReference Include="path/to/Kode.Agent.Store.Json.csproj" />
<ProjectReference Include="path/to/Kode.Agent.Tools.Builtin.csproj" />

2. Configure Environment Variables

Create a .env file (Anthropic Claude recommended):

# Anthropic configuration (recommended)
ANTHROPIC_API_KEY=your-api-key-here
ANTHROPIC_MODEL_ID=claude-sonnet-4-20250514
ANTHROPIC_BASE_URL=https://api.anthropic.com  # Optional, supports custom endpoint

Optional configuration for other providers:

# OpenAI configuration (optional)
OPENAI_API_KEY=your-api-key-here
OPENAI_MODEL_ID=gpt-5.2
OPENAI_BASE_URL=https://api.openai.com  # Optional, supports Azure OpenAI

# Zhipu AI (optional)
DEFAULT_PROVIDER=anthropic
ANTHROPIC_API_KEY=your-api-key-here
ANTHROPIC_MODEL_ID=GLM-4.7
ANTHROPIC_BASE_URL=https://open.bigmodel.cn/api/anthropic

Tip: You can also configure API keys directly in code without a .env file.

3. Basic Usage

using Kode.Agent.Sdk.Core.Abstractions;
using Kode.Agent.Sdk.Core.Types;
using Kode.Agent.Sdk.Infrastructure.Providers;
using Kode.Agent.Sdk.Tools;
using Kode.Agent.Store.Json;
using Kode.Agent.Tools.Builtin;
using AgentImpl = Kode.Agent.Sdk.Core.Agent.Agent;

// 1. Create storage
var store = new JsonAgentStore("./.kode");

// 2. Register tools
var toolRegistry = new ToolRegistry();
toolRegistry.RegisterBuiltinTools();

// 3. Create model provider (Anthropic Claude recommended)
var provider = new AnthropicProvider(new AnthropicOptions
{
    ApiKey = Environment.GetEnvironmentVariable("ANTHROPIC_API_KEY")!,
    BaseUrl = Environment.GetEnvironmentVariable("ANTHROPIC_BASE_URL")
});

// 4. Create dependencies
var deps = new AgentDependencies
{
    Store = store,
    ToolRegistry = toolRegistry,
    ModelProvider = provider,
    SandboxFactory = new LocalSandboxFactory()
};

// 5. Configure agent
var config = new AgentConfig
{
    Model = "claude-sonnet-4-20250514",
    SystemPrompt = "You are a helpful assistant.",
    MaxIterations = 10,
    Tools = ["fs_read", "fs_glob", "fs_grep", "shell_exec"]
};

// 6. Create and run agent
var agent = await AgentImpl.CreateAsync("my-agent", config, deps);
await agent.RunAsync("Help me analyze the project structure of the current directory");

Event Subscriptions

Agent provides three event channels for different scenarios:

graph LR
    subgraph Agent["Agent"]
        EventBus[EventBus]
    end

    subgraph Channels["Event Channels"]
        Progress["Progress<br/>Real-time Output Stream"]
        Control["Control<br/>Approval Control"]
        Monitor["Monitor<br/>Logging Monitor"]
    end

    subgraph Handlers["Handlers"]
        UI["UI Rendering"]
        Approval["Approval Handling"]
        Logger["Logging"]
    end

    EventBus --> Progress --> UI
    EventBus --> Control --> Approval
    EventBus --> Monitor --> Logger

    style Progress fill:#e8f5e9
    style Control fill:#fff3e0
    style Monitor fill:#e3f2fd

// Progress channel - Real-time output stream
await foreach (var envelope in agent.EventBus.SubscribeAsync(EventChannel.Progress))
{
    switch (envelope.Event)
    {
        case TextChunkEvent textChunk:
            Console.Write(textChunk.Delta);  // Stream text output
            break;

        case ToolStartEvent toolStart:
            Console.WriteLine($"[tool] {toolStart.Call.Name} starting...");
            break;

        case ToolEndEvent toolEnd:
            Console.WriteLine($"[tool] {toolEnd.Call.Name} completed");
            break;

        case DoneEvent:
            Console.WriteLine("Conversation complete");
            break;
    }
}

// Control channel - Events requiring human intervention
await foreach (var envelope in agent.EventBus.SubscribeAsync(EventChannel.Control))
{
    if (envelope.Event is PermissionRequiredEvent approval)
    {
        // Handle operations requiring approval (example: allow all)
        await agent.ApproveToolCallAsync(approval.Call.Id);
    }
}

// Monitor channel - Logging and debugging information
await foreach (var envelope in agent.EventBus.SubscribeAsync(EventChannel.Monitor))
{
    if (envelope.Event is ErrorEvent err)
    {
        Console.WriteLine($"[{err.Severity}/{err.Phase}] {err.Message}");
    }
}

Custom Tools

Method 1: Using Source Generator (Recommended)

using Kode.Agent.Sdk.Tools;

[Tool("weather")]
[Description("Get current weather for a location")]
public partial class WeatherTool : ITool
{
    [ToolParameter("location", required: true)]
    [Description("City name or coordinates")]
    public string Location { get; set; } = "";

    [ToolParameter("unit")]
    [Description("Temperature unit: celsius or fahrenheit")]
    public string Unit { get; set; } = "celsius";

    public async Task<ToolResult> ExecuteAsync(ToolContext context)
    {
        // Implement weather fetching logic
        var weather = await FetchWeather(Location, Unit);
        return ToolResult.Success(weather);
    }
}

Method 2: Manual Registration

toolRegistry.Register(new ToolDefinition
{
    Name = "calculator",
    Description = "Perform mathematical calculations",
    InputSchema = new
    {
        type = "object",
        properties = new
        {
            expression = new { type = "string", description = "Math expression to evaluate" }
        },
        required = new[] { "expression" }
    }
}, async (input, context) =>
{
    var expr = input["expression"]?.ToString();
    var result = EvaluateExpression(expr);
    return ToolResult.Success(result.ToString());
});

Model Providers

Anthropic Claude

var provider = new AnthropicProvider(new AnthropicOptions
{
    ApiKey = "your-api-key",
    BaseUrl = "https://api.anthropic.com",  // Or custom endpoint
    ModelId = "claude-sonnet-4-20250514",
    EnableBetaFeatures = false
});

Supported models:

claude-sonnet-4-20250514
claude-3-5-sonnet-20241022
claude-3-5-haiku-20241022
claude-3-opus-20240229

OpenAI GPT

var provider = new OpenAIProvider(new OpenAIOptions
{
    ApiKey = "your-api-key",
    BaseUrl = "https://api.openai.com",  // Or Azure OpenAI endpoint
    Organization = "your-org-id",  // Optional
    DefaultModel = "gpt-4o"
});

Supported models:

gpt-4o
gpt-4o-mini
gpt-4-turbo
o1-preview
o1-mini

Using OpenAI Compatible Endpoints

Many third-party services provide OpenAI-compatible APIs:

// Zhipu AI (BigModel)
var provider = new AnthropicProvider(new AnthropicOptions
{
    ApiKey = "your-key",
    BaseUrl = "https://open.bigmodel.cn/api/anthropic"
});

// DeepSeek
var provider = new OpenAIProvider(new OpenAIOptions
{
    ApiKey = "your-key",
    BaseUrl = "https://api.deepseek.com",
    DefaultModel = "deepseek-chat"
});

Dependency Injection

Using Microsoft.Extensions.DependencyInjection:

using Kode.Agent.Sdk.Extensions;

var services = new ServiceCollection();

// Register Agent SDK
services.AddKodeAgent(options =>
{
    options.DefaultModel = "claude-sonnet-4-20250514";
    options.StoreDirectory = "./.kode";
});

// Register Anthropic provider (recommended)
services.AddAnthropicProvider(options =>
{
    options.ApiKey = Configuration["Anthropic:ApiKey"]!;
    options.BaseUrl = Configuration["Anthropic:BaseUrl"];
});

// Or register OpenAI provider (optional)
// services.AddOpenAIProvider(options =>
// {
//     options.ApiKey = Configuration["OpenAI:ApiKey"]!;
// });

var serviceProvider = services.BuildServiceProvider();

// Use
var agentFactory = serviceProvider.GetRequiredService<IAgentFactory>();
var agent = await agentFactory.CreateAsync("my-agent", config);

MCP Integration

SDK has native support for Model Context Protocol (MCP), making it easy to integrate external tool ecosystems:

using Kode.Agent.Mcp;

// Configure MCP servers
var mcpConfig = new McpConfig
{
    Servers = new Dictionary<string, McpServerConfig>
    {
        ["filesystem"] = new()
        {
            Command = "npx",
            Args = ["-y", "@modelcontextprotocol/server-filesystem", "/tmp"]
        },
        ["github"] = new()
        {
            Command = "npx",
            Args = ["-y", "@modelcontextprotocol/server-github"],
            Env = new() { ["GITHUB_TOKEN"] = Environment.GetEnvironmentVariable("GITHUB_TOKEN")! }
        }
    }
};

// Register MCP tools
services.AddMcpTools(mcpConfig);

Permission Control

SDK provides a flexible permission control mechanism:

var config = new AgentConfig
{
    Model = "claude-sonnet-4-20250514",
    Tools = ["fs_read", "fs_write", "bash_run"],
    Permissions = new PermissionConfig
    {
        Mode = "auto",                       // auto | approval | readonly | custom
        RequireApprovalTools = ["bash_run"], // Tools requiring approval
        DenyTools = ["fs_rm"]                // Forbidden tools
    }
};

Permission modes:

auto - Default allow (can be refined with RequireApprovalTools/DenyTools/AllowTools)
approval - All tools require manual approval (returns permission_required)
readonly - Based on ToolDescriptor.metadata to determine if it "has side effects", tools that mutate state are directly denied, others allow/ask (aligned with TS permission-modes)
Custom mode - Requires registering corresponding permission mode handler in host process

Built-in Tools

SDK provides a rich set of built-in tools:

File System Tools

Tool Name	Description	Permission
`fs_read`	Read file content, supports line ranges	Read-only
`fs_write`	Write file content	Requires approval
`fs_edit`	Edit file (search and replace)	Requires approval
`fs_multi_edit`	Batch edit multiple files	Requires approval
`fs_glob`	File pattern search	Read-only
`fs_grep`	File content search (regex)	Read-only
`fs_list`	List directory contents	Read-only
`fs_rm`	Delete file or directory	Requires approval

Shell Tools

Tool Name	Description	Permission
`bash_run`	Execute Shell commands	Requires approval
`bash_logs`	View command execution logs	Read-only
`bash_kill`	Terminate background process	Requires approval

Task Management Tools

Tool Name	Description	Permission
`todo_read`	Read Todo list	Read-only
`todo_write`	Update Todo list	Requires approval
`task_run`	Run predefined tasks	Requires approval

Skills Tools

Tool Name	Description	Permission
`skill_list`	List available skills	Read-only
`skill_activate`	Activate skill	Requires approval
`skill_resource`	Get skill resources	Read-only

State Persistence

// Continuously persists to Store during execution (messages/tool-calls/todos/meta/events)
await agent.RunAsync("Start task");

// Save a "safe fork point" snapshot (TS-aligned: agent.snapshot)
var snapshotId = await agent.SnapshotAsync();

// Resume from Store (TS-aligned: resumeFromStore)
var restoredAgent = await Agent.ResumeFromStoreAsync("agent-id", deps);
await restoredAgent.RunAsync("Continue our previous discussion");

Project Structure

csharp/
├── src/
│   ├── Kode.Agent.Sdk/              # Core SDK
│   │   ├── Core/                    # Core abstractions and types
│   │   │   ├── Abstractions/        # IAgent, IEventBus, ITool etc.
│   │   │   ├── Agent/               # Agent implementation
│   │   │   ├── Events/              # Event system
│   │   │   ├── Types/               # AgentConfig, Message etc.
│   │   │   └── Todo/                # Todo management
│   │   ├── Infrastructure/          # Infrastructure
│   │   │   ├── Providers/           # Anthropic/OpenAI providers
│   │   │   └── Sandbox/             # Sandbox execution
│   │   ├── Tools/                   # Tool system (ToolRegistry, ToolBase)
│   │   └── Extensions/              # DI extensions
│   ├── Kode.Agent.SourceGenerator/  # Compile-time code generator
│   ├── Kode.Agent.Mcp/              # MCP protocol support
│   ├── Kode.Agent.Store.Json/       # JSON file storage
│   ├── Kode.Agent.Store.Redis/      # Redis distributed storage
│   └── Kode.Agent.Tools.Builtin/    # Built-in tool implementations
├── examples/
│   ├── Kode.Agent.Examples/         # Usage examples (Console)
│   └── Kode.Agent.WebApiAssistant/  # Usage examples (ASP.NET WebAPI, OpenAI SSE compatible)
├── tests/
│   └── Kode.Agent.Tests/            # Unit and integration tests
└── docs/                            # Detailed documentation
    ├── ADVANCED_GUIDE.md            # Advanced guide
    └── API_REFERENCE.md             # API reference

Requirements

.NET 10.0 or higher
Valid Anthropic or OpenAI API key

Documentation

Core Documentation

Document	Description
Advanced Guide	Deep dive into architecture design, event system, tool development, Skills system, Sub-Agent delegation, MCP integration, permission control, state management, and other advanced topics
API Reference	Complete API documentation, including core types (AgentConfig, AgentDependencies), Agent lifecycle, event models, MCP integration types, Store interfaces, etc.
TypeScript Alignment	C# SDK and TypeScript implementation alignment status, difference analysis, and continuous alignment plan

Quick Navigation

Architecture

Overall Architecture Overview - SDK components, dependencies, core component descriptions
Event System Details - Three-channel architecture, event types, subscription patterns
Agent Lifecycle - State transitions, breakpoint states, creation and recovery

Tool Development

Tool Development Guide - Tool interfaces, Source Generator, registration methods
MCP Protocol Integration - MCP configuration, common servers, tool filtering
Built-in Tools List - File system, Shell, task management, skills tools

Advanced Features

Skills System - Progressive skill discovery and activation
Sub-Agent Delegation - Task decomposition and template management
Permission Control System - Permission modes, configuration examples, approval process
State Storage - JSON/Redis storage, resumable execution

Best Practices

Error Handling - Exception types, handling patterns
Best Practices - Serilog logging, timeout settings, resource management, session management

Running Examples

Console Examples

cd examples/Kode.Agent.Examples

# Copy environment variable template
cp .env.example .env
# Edit .env with your API keys

# Run examples
dotnet run

Available examples:

GettingStarted - Basic usage, quick start
AgentInbox - Event flow and tool execution monitoring
ApprovalControl - Manual approval flow control
RoomCollab - Multi-agent collaboration scenarios
CustomToolsExample - Custom tool development
HooksUsage - Lifecycle hooks usage
TemplateUsage - Agent template system
SchedulerUsage - Scheduled task management
EventBusUsage - Event bus details

WebAPI Example (OpenAI Compatible)

This is an ASP.NET WebAPI application that exposes an OpenAI Chat Completions compatible interface with SSE streaming support.

Special Offer: Zhipu GLM Coding value subscription, supporting Claude Code, Cline, and 20+ major programming tools!

Join now for limited-time special pricing!

Quick Start

cd examples/Kode.Agent.WebApiAssistant

cp .env.example .env
# Edit .env with your API keys

dotnet run

Default listening address is shown in console output (usually http://localhost:5123).

Using OpenAI Client

Since the interface is fully OpenAI compatible, you can use any client that supports the OpenAI API:

Configuration URL: http://localhost:5123/v1/chat/completions

Example - Using curl:

# Non-streaming request
curl http://localhost:5123/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-4-20250514",
    "user": "my-session",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "Hello, introduce yourself"}
    ],
    "stream": false
  }'

# Streaming request (SSE)
curl http://localhost:5123/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Accept: text/event-stream" \
  -d '{
    "model": "claude-sonnet-4-20250514",
    "user": "my-session",
    "messages": [
      {"role": "user", "content": "Summarize today's plan in 3 sentences"}
    ],
    "stream": true
  }'

Supported Clients:

OpenAI Python SDK
OpenAI Node.js SDK
LangChain
LlamaIndex
Cursor
Continue
Cline
Any other tool that supports custom OpenAI endpoints

Python Example:

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:5123/v1",
    api_key="anything"  # No real key needed
)

response = client.chat.completions.create(
    model="claude-sonnet-4-20250514",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Help me analyze this project"}
    ],
    stream=True
)

for chunk in response:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")

Node.js Example:

import OpenAI from 'openai';

const openai = new OpenAI({
  baseURL: 'http://localhost:5123/v1',
  apiKey: 'anything'  // No real key needed
});

const stream = await openai.chat.completions.create({
  model: 'claude-sonnet-4-20250514',
  messages: [
    { role: 'system', content: 'You are a helpful assistant.' },
    { role: 'user', content: 'Help me analyze this project' }
  ],
  stream: true
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content || '');
}

Key Features

Feature	Description
OpenAI Compatible	Fully compatible with OpenAI Chat Completions API
SSE Streaming	Supports `text/event-stream` streaming responses
Session Persistence	Automatically saves conversation state, supports multi-turn conversations
Tool Calling	Supports file system, shell, email, notification, and other tools
Permission Control	Fine-grained tool permission management
MCP Integration	Supports Model Context Protocol external tools

For more detailed interface documentation, see examples/Kode.Agent.WebApiAssistant/README.md.

License

MIT License

Contributing

Issues and Pull Requests are welcome! Please check the Contributing Guide for more information.

Community

Join our WeChat group to discuss technical questions and best practices:

Name		Name	Last commit message	Last commit date
Latest commit History 19 Commits
.github		.github
assets		assets
docs		docs
examples		examples
src		src
tests		tests
.gitignore		.gitignore
CLAUDE.md		CLAUDE.md
Directory.Build.props		Directory.Build.props
Directory.Packages.props		Directory.Packages.props
Kode.Agent.sln		Kode.Agent.sln
LICENSE		LICENSE
NuGet.Config		NuGet.Config
PUBLISH_GUIDE.md		PUBLISH_GUIDE.md
README-zh.md		README-zh.md
README.md		README.md
global.json		global.json
nuget.config.github.example		nuget.config.github.example

Link	Description
Kode SDK (TypeScript)	TypeScript version of the SDK with feature parity
Example Projects	Console and WebAPI examples
Source Code	Core SDK implementation code
Test Code	Unit and integration tests
Changelog	Version update records

License

featbit/featbit-front-agent-api

Folders and files

Latest commit

History

Repository files navigation

Kode Agent SDK for .NET

Project Background

Table of Contents

Features

Architecture Overview

Agent Execution Flow

Quick Start

1. Install Packages

2. Configure Environment Variables

3. Basic Usage

Event Subscriptions

Custom Tools

Method 1: Using Source Generator (Recommended)

Method 2: Manual Registration

Model Providers

Anthropic Claude

OpenAI GPT

Using OpenAI Compatible Endpoints

Dependency Injection

MCP Integration

Permission Control

Built-in Tools

File System Tools

Shell Tools

Task Management Tools

Skills Tools

State Persistence

Project Structure

Requirements

Documentation

Core Documentation

Quick Navigation

Architecture

Tool Development

Advanced Features

Best Practices

Running Examples

Console Examples

WebAPI Example (OpenAI Compatible)

Quick Start

Using OpenAI Client

Key Features

License

Contributing

Community

Related Links

Ecosystem

About

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Languages

Packages