create-agent-provider — microsoft-agent-framework create-agent-provider, AgentFrameworkToolkit, community, microsoft-agent-framework, ide skills, Claude Code, Cursor, Windsurf

v1.0.0
GitHub

About this Skill

Ideal for AI Agents like Cursor, Windsurf, and Claude Code needing seamless integration with various LLM services through the Microsoft Agent Framework An opinionated C# Toolkit on top of Microsoft Agent Framework that makes life easier

# Core Topics

rwjdk rwjdk
[53]
[5]
Updated: 2/28/2026

Agent Capability Analysis

The create-agent-provider skill by rwjdk is an open-source community AI agent skill for Claude Code and other IDE workflows, helping agents execute tasks with better context, repeatability, and domain-specific guidance. Optimized for microsoft-agent-framework.

Ideal Agent Persona

Ideal for AI Agents like Cursor, Windsurf, and Claude Code needing seamless integration with various LLM services through the Microsoft Agent Framework

Core Value

Empowers agents to integrate with different LLM services such as Anthropic, OpenAI, and Google using the AgentFrameworkToolkit, enabling the creation of custom Agent Provider NuGet packages for expanded functionality and compatibility with OpenAI-compatible APIs

Capabilities Granted for create-agent-provider

Creating custom Agent Providers for unsupported LLM services
Integrating OpenAI-compatible LLM services like OpenRouter and XAI with the AgentFrameworkToolkit
Developing NuGet packages for easy distribution and deployment of custom Agent Providers

! Prerequisites & Limits

  • Requires Microsoft Agent Framework and C# expertise
  • Limited to LLM services with OpenAI-compatible APIs or custom implementations
Labs Demo

Browser Sandbox Environment

⚡️ Ready to unleash?

Experience this Agent in a zero-setup browser environment powered by WebContainers. No installation required.

Boot Container Sandbox

create-agent-provider

Install create-agent-provider, an AI agent skill for AI agent workflows and automation. Works with Claude Code, Cursor, and Windsurf with one-command setup.

SKILL.md
Readonly

Create Agent Provider Package

This skill guides you through creating a new Agent Provider NuGet package for the AgentFrameworkToolkit. Agent Providers enable integration with different LLM services (e.g., Anthropic, OpenAI, Google, Mistral).

Quick Decision Guide

Choose implementation approach:

  1. OpenAI-Compatible Provider - If the LLM service has an OpenAI-compatible API:

  2. Custom Provider - If the LLM service has a unique API:

    • Examples: Anthropic (Claude), Google (Gemini), GitHub Models, Mistral
    • See: Custom Provider Template
    • Full control, follows standard patterns

Core Provider Components

Most provider packages include these components (OpenAI-compatible providers reuse AgentFrameworkToolkit.OpenAI.AgentOptions):

1. Connection Class (<Provider>Connection)

  • Manages API credentials and configuration
  • Creates and configures the SDK client
  • Supports network timeout and custom endpoints (if applicable)
  • Provides raw HTTP call inspection hooks (if possible)

2. Agent Factory (<Provider>AgentFactory)

  • Creates agent instances from options (or from simplified overloads)
  • Builds the inner ChatClientAgent (custom providers)
  • Applies middleware via AgentFrameworkToolkit.MiddlewareHelper

3. Agent Options (<Provider>AgentOptions) (custom providers only)

  • Configuration for agent creation (model, instructions, tools, max tokens, etc.)
  • Provider-specific settings (e.g., thinking budget)
  • Middleware configuration (raw HTTP/tool inspection, tool-calling middleware, OpenTelemetry, logging)

4. Agent Wrapper (<Provider>Agent)

  • Inherits from AIAgent
  • Delegates to an inner agent instance
  • Exposes an InnerAgent property

5. Chat Models Constants (<Provider>ChatModels) (only if provider is specific. if it offers multiple LLMs don't include such)

  • Constants for available model IDs
  • Makes model selection discoverable

Implementation Steps

Step 1: Project Setup

bash
1mkdir src/AgentFrameworkToolkit.<Provider>

Important repo conventions (don’t fight the build system):

  • src/* projects inherit defaults from Directory.Build.props (currently net8.0).
  • NuGet packaging defaults are imported via nuget-package.props in each src/* .csproj.
  • Central package versions are in Directory.Packages.props (no versions in .csproj).

Step 2: Implement Core Components

Follow existing providers:

  • OpenAI-compatible: src/AgentFrameworkToolkit.OpenRouter/, src/AgentFrameworkToolkit.XAI/, src/AgentFrameworkToolkit.Cohere/
  • Custom: src/AgentFrameworkToolkit.Anthropic/, src/AgentFrameworkToolkit.GitHub/, src/AgentFrameworkToolkit.Google/, src/AgentFrameworkToolkit.Mistral/

Step 3: Add Service Extensions

Create ServiceCollectionExtensions.cs for dependency injection using the repo naming convention:

  • Add<Provider>AgentFactory(this IServiceCollection services, string apiKey)
  • Add<Provider>AgentFactory(this IServiceCollection services, <Provider>Connection connection)

Step 4: Write Tests (Integration Tests)

Provider tests live in development/Tests/ and make real API calls. See Testing Guide for the repo’s concrete pattern.

Quick checklist:

  1. Add provider project references:
    • development/Tests/Tests.csproj
    • development/Sandbox/Sandbox.csproj
  2. Add provider to the shared test harness:
    • Add a value to AgentProvider enum in development/Tests/TestBase.cs
    • Add a case to GetAgentForScenarioAsync(...) in development/Tests/TestBase.cs
  3. Create development/Tests/<Provider>Tests.cs:
    • Call the shared scenario tests (SimpleAgentTestsAsync, NormalAgentTestsAsync, etc.)
    • Add DI tests for Add<Provider>AgentFactory(...)
  4. Add a sandbox runner in development/Sandbox/Providers/<Provider>.cs and optionally wire it in development/Sandbox/Program.cs
  5. Add your API key to user-secrets:
    • Update development/Secrets/Secrets.cs and development/Secrets/SecretsManager.cs
    • Set the secret using dotnet user-secrets for the development/Secrets/Secrets.csproj project

Step 5: Repository Integration

  1. Add the project to AgentFrameworkToolkit.slnx under /Packages/
  2. Add any new SDK package versions to Directory.Packages.props
  3. Update documentation:
    • Main README.md provider table
    • Provider-specific src/AgentFrameworkToolkit.<Provider>/README.md
    • CHANGELOG.md

Step 6: Validation

bash
1dotnet build --configuration Release

Key Architectural Patterns

Middleware Configuration (custom providers)

Use the shared helper instead of re-implementing ordering rules:

  • AgentFrameworkToolkit.MiddlewareHelper.ApplyMiddleware(...)

Agent Factory Pattern (custom providers)

csharp
1public class <Provider>AgentFactory 2{ 3 public <Provider>Connection Connection { get; } 4 5 public <Provider>AgentFactory(string apiKey); 6 public <Provider>AgentFactory(<Provider>Connection connection); 7 8 public <Provider>Agent CreateAgent(<Provider>AgentOptions options) 9 { 10 // 1. Get SDK client from connection 11 // 2. Build IChatClient (or use SDK-provided one) 12 // 3. Create ChatClientAgent 13 // 4. Apply middleware via MiddlewareHelper 14 // 5. Wrap in provider-specific agent 15 } 16}

Common Pitfalls

  1. Missing XML documentation (warnings are errors in this repo)
  2. Putting versions in .csproj instead of Directory.Packages.props
  3. Forgetting to update AgentFrameworkToolkit.slnx and README.md
  4. Forgetting to add provider to development/Tests/TestBase.cs test harness
  5. Hardcoding API keys instead of using user-secrets (development/Secrets/SecretsManager.cs)

Reference Documentation

FAQ & Installation Steps

These questions and steps mirror the structured data on this page for better search understanding.

? Frequently Asked Questions

What is create-agent-provider?

Ideal for AI Agents like Cursor, Windsurf, and Claude Code needing seamless integration with various LLM services through the Microsoft Agent Framework An opinionated C# Toolkit on top of Microsoft Agent Framework that makes life easier

How do I install create-agent-provider?

Run the command: npx killer-skills add rwjdk/AgentFrameworkToolkit/create-agent-provider. It works with Cursor, Windsurf, VS Code, Claude Code, and 19+ other IDEs.

What are the use cases for create-agent-provider?

Key use cases include: Creating custom Agent Providers for unsupported LLM services, Integrating OpenAI-compatible LLM services like OpenRouter and XAI with the AgentFrameworkToolkit, Developing NuGet packages for easy distribution and deployment of custom Agent Providers.

Which IDEs are compatible with create-agent-provider?

This skill is compatible with Cursor, Windsurf, VS Code, Trae, Claude Code, OpenClaw, Aider, Codex, OpenCode, Goose, Cline, Roo Code, Kiro, Augment Code, Continue, GitHub Copilot, Sourcegraph Cody, and Amazon Q Developer. Use the Killer-Skills CLI for universal one-command installation.

Are there any limitations for create-agent-provider?

Requires Microsoft Agent Framework and C# expertise. Limited to LLM services with OpenAI-compatible APIs or custom implementations.

How To Install

  1. 1. Open your terminal

    Open the terminal or command line in your project directory.

  2. 2. Run the install command

    Run: npx killer-skills add rwjdk/AgentFrameworkToolkit/create-agent-provider. The CLI will automatically detect your IDE or AI agent and configure the skill.

  3. 3. Start using the skill

    The skill is now active. Your AI agent can use create-agent-provider immediately in the current project.

Related Skills

Looking for an alternative to create-agent-provider or another community skill for your workflow? Explore these related open-source skills.

View All

widget-generator

Logo of f
f

f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.

149.6k
0
AI

flags

Logo of vercel
vercel

flags is a Next.js feature management skill that enables developers to efficiently add or modify framework feature flags, streamlining React application development.

138.4k
0
Browser

zustand

Logo of lobehub
lobehub

The ultimate space for work and life — to find, build, and collaborate with agent teammates that grow with you. We are taking agent harness to the next level — enabling multi-agent collaboration, effortless agent team design, and introducing agents as the unit of work interaction.

72.8k
0
AI

data-fetching

Logo of lobehub
lobehub

The ultimate space for work and life — to find, build, and collaborate with agent teammates that grow with you. We are taking agent harness to the next level — enabling multi-agent collaboration, effortless agent team design, and introducing agents as the unit of work interaction.

72.8k
0
AI