Skip to content

README.md

Latest commit

 

History

History
91 lines (60 loc) · 2.68 KB

README.md

File metadata and controls

91 lines (60 loc) · 2.68 KB

ArbiterAI Examples

This directory contains example applications demonstrating various features of the ArbiterAI library.

Available Examples

CLI Chat Client (cli/)

A command-line interface for interactive chat sessions with LLM models.

  • Interactive multi-turn chat
  • Multiple providers (OpenAI, Anthropic, DeepSeek, OpenRouter, local models)
  • Model selection and configuration

See cli/main.cpp for details.

Build target: arbiterAI-cli

HTTP Proxy Server (proxy/)

An OpenAI-compatible HTTP proxy server demonstrating how to build production services with ArbiterAI.

Build target: arbiterAI-proxy

Mock Provider Example (mock_example.cpp)

Demonstrates the Mock provider for testing without actual LLM calls. Shows echo tag usage, streaming, conversation history, and usage statistics.

Note: This example is not currently wired into CMakeLists.txt as a build target.

Mock Provider Usage

The Mock provider enables testing without network calls or API keys. Use <echo> tags to control responses:

#include "arbiterAI/arbiterAI.h"
#include "arbiterAI/chatClient.h"

// Initialize with mock configuration
arbiterAI::ArbiterAI& ai = arbiterAI::ArbiterAI::instance();
ai.initialize({"examples/mock_models.json"});

// Create chat client with mock model
arbiterAI::ChatConfig config;
config.model = "mock-model";
auto client = ai.createChatClient(config);

// Use echo tags to control response
arbiterAI::CompletionRequest request;
request.messages = {{"user", "What is 2+2? <echo>4</echo>"}};

arbiterAI::CompletionResponse response;
client->completion(request, response);
// response.text == "4"

Echo Tag Quick Reference

Pattern Result
<echo>text</echo> "text"
<echo>\nline1\nline2\n</echo> "line1\nline2"
<ECHO>text</ECHO> "text" (case insensitive)
Multiple tags First match is used
No tag Default mock message

See the Testing Guide for comprehensive documentation.

Configuration Files

File Description
mock_models.json Mock model definitions for testing (no API keys needed)
model_config_v2.json Production model configurations for real LLM providers

Building

All examples are built as part of the main ArbiterAI build:

# Inside Docker container
./build.sh

Build output is in build/linux_x64_debug/.

Further Documentation