nexus-ai-cli 0.4.0

Ultra-minimal multi-provider AI framework

Nexus AI

Ultra-minimal multi-provider AI CLI framework.

Available on PyPI as nexus-ai-cli

Features

• Unified Interface: Single CLI for OpenAI, Anthropic, Ollama, and OpenRouter
• Session Management: Persistent conversations with search and export
• Prompt Library: Save and reuse system prompts
• Smart Defaults: Interactive model selection and fuzzy search
• Streaming Output: Real-time response streaming

Installation

Quick Install (Recommended)

pipx install nexus-ai-cli

pipx creates an isolated environment for Nexus AI, preventing conflicts with other Python packages.

Alternative Installation Methods

Using pip:

pip install nexus-ai-cli

Development install:

git clone https://github.com/saadiqhorton/nexus-ai.git
cd nexus-ai
pip install -e .

Updating

pipx upgrade nexus-ai-cli
# or
pip install --upgrade nexus-ai-cli

System Requirements

• Python 3.8 or higher
• API keys for at least one AI provider (OpenAI, Anthropic, or OpenRouter)
• Ollama users: Ollama server running at http://localhost:11434

Environment Setup

Copy the example environment file and add your API keys:

cp .env.example .env

Required environment variables (set at least one provider):

Variable	Description
OPENAI_API_KEY	OpenAI API key
ANTHROPIC_API_KEY	Anthropic API key
OPENROUTER_API_KEY	OpenRouter API key

For local Ollama, no API key is needed - just ensure the server is running at http://localhost:11434.

Quick Start

# Use default model
nexus "your prompt"

# Specify model
nexus -m gpt-4o "your prompt"

# Include file contents
nexus -f file.txt "explain this"

# Pipe content
cat file.txt | nexus "summarize"

# Interactive chat
nexus chat --session myproject

Commands

Command	Description
nexus "prompt"	Send a prompt
nexus chat	Interactive REPL
nexus models	List available models
nexus providers	Show configured providers
nexus config	Display configuration
nexus -d [model]	Set default model
nexus sessions	Manage conversations
nexus prompts	Manage prompt library
nexus completion	Setup shell completion

Configuration

Config stored at ~/.nexus/config.yaml:

providers:
  openai:
    api_key: ${OPENAI_API_KEY}
  anthropic:
    api_key: ${ANTHROPIC_API_KEY}
  ollama:
    base_url: http://localhost:11434
  openrouter:
    api_key: ${OPENROUTER_API_KEY}

defaults:
  provider: openai
  model: gpt-4o
  temperature: 0.7
  max_tokens: 2000
  stream: true

Development

# Install with dev dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Run tests with Ollama integration
NEXUS_OLLAMA_URL=http://localhost:11434 pytest

# Linting
ruff check .
ruff format .

Architecture

nexus/
├── cli/           # Click CLI commands
├── core/          # App, ProviderManager, CompletionHandler
├── providers/     # AI provider implementations
├── config/        # ConfigManager and models
├── session/       # Session persistence
├── prompts/       # Prompt library
└── utils/         # Cache, logging, errors