AgentDock Core Documentation

Provider-Agnostic API Architecture

This document explains how AgentDock works with multiple LLM providers through a unified API layer.

What This Means For Users

AgentDock supports a variety of LLM providers (OpenAI, Anthropic, Gemini, DeepSeek, Groq, etc.) through a single, consistent interface. This means:

  1. Consistent Experience - The same chat interface works across all providers
  2. Easy Provider Switching - Change providers without changing your application code
  3. Unified Error Handling - Clear, consistent error messages regardless of provider
  4. Simple API Key Management - Manage all provider keys in one place

Benefits

For Developers

  1. Simplified Integration - Connect to any supported provider using the same API
  2. No Provider-Specific Code - Write code once that works with all providers
  3. Future-Proof - New providers are added to the core library without requiring changes to your application
  4. Type Safety - Full TypeScript support for all providers

For End Users

  1. Provider Flexibility - Use preferred providers without learning new interfaces
  2. Graceful Error Handling - Receive clear, actionable error messages
  3. Consistent Model Selection - Choose models through a standardized interface
  4. Smooth Failover - Automatic retries and provider fallbacks when configured

Supported Providers

AgentDock currently supports these LLM providers:

  • OpenAI - GPT models (3.5, 4, etc.)
  • Anthropic - Claude models
  • Google - Gemini models
  • DeepSeek - DeepSeek models
  • Groq - Fast inference for various models

Error Handling

The provider-agnostic design includes standardized error handling:

  1. Normalized Errors - Technical provider errors are translated to user-friendly messages
  2. Clear API Key Guidance - Specific instructions when API keys are missing or invalid
  3. Appropriate Recovery Options - Context-aware options for resolving different error types

How It Works

Behind the scenes, AgentDock:

  1. Accepts a standard message format for all providers
  2. Translates requests to provider-specific formats
  3. Manages streaming connections appropriately for each provider
  4. Normalizes responses and errors back to a standard format

This abstraction layer means that applications using AgentDock don't need to know the details of each provider's API.