Welcome to LLxprt Code documentation

This documentation provides a comprehensive guide to installing, using, and developing LLxprt Code. This tool lets you interact with AI models from multiple providers through a command-line interface.

Overview

LLxprt Code brings the capabilities of large language models to your terminal in an interactive Read-Eval-Print Loop (REPL) environment. LLxprt Code consists of a client-side application (packages/cli) that communicates with a local server (packages/core), which in turn manages requests to the provider APIs and its AI models. LLxprt Code also contains a variety of tools for tasks such as performing file system operations, running shells, and web fetching, which are managed by packages/core.

Navigating the documentation

Getting Started

New to LLxprt Code? Start here:

• Getting Started Guide: Your first session — installation, authentication, and your first AI-assisted coding task.

Core Workflows

Essential guides for daily use:

• Provider Configuration: Configure and manage LLM providers (Anthropic, OpenAI, Gemini, and more).
• Authentication: Set up authentication for various providers.
• Profiles: Save and manage configurations for different contexts.
• Sandboxing: Protect your system with container-based isolation.
• Subagents: Create and manage specialized AI assistants tied to profiles.

Advanced Topics

Power features for advanced users:

• OAuth Setup: Configure OAuth authentication for subscription access.
• Local Models: Complete guide to using local AI models with LM Studio, Ollama, llama.cpp, and more.
• Zed Editor Integration: Connect LLxprt Code to the Zed editor as an AI assistant.
• OpenAI Responses API: Using OpenAI's enhanced Responses API.
• Prompt Configuration: How to customize AI behavior with prompts.
• Settings and Profiles: How to manage settings and use profiles.
• Checkpointing: Save and restore session state.
• Extensions: How to extend the CLI with new functionality.
• IDE Integration: Connect the CLI to your editor.

Reference

Configuration, commands, and troubleshooting:

• Configuration: Configure the CLI behavior.
• Commands Reference: Complete documentation of available CLI commands.
• Troubleshooting Guide: Find solutions to common problems and FAQs.
• CLI Introduction: Overview of the command-line interface.
• Execution and Deployment: Information for running LLxprt Code.
• Keyboard Shortcuts: Quick reference for keyboard shortcuts.
• Themes: Customize the visual appearance.
• Emoji Filter: Control emoji usage in LLM responses.
• Runtime helper APIs: Reference for runtime-scoped helper functions.
• Context Dumping: Capture API requests/responses for debugging and curl replay.
• Telemetry: Overview of telemetry in the CLI.
• Telemetry Privacy: Privacy information for telemetry.
• Migration from Gemini CLI: Guide for users migrating from Gemini CLI.

Architecture & Development

For contributors and developers:

• Architecture Overview: Understand the high-level design of LLxprt Code.
• Core Introduction: Overview of the core component.
• Provider runtime context: Manage ProviderRuntimeContext lifecycles.
• Provider interface: Implement providers against the stateless runtime.
• Tools API: Information on how the core manages and exposes tools.
• Memory Import Processor: Modular LLXPRT.md import feature.
• Shell Replacement: Command substitution in shell commands.
• Contributing & Development Guide: Information for contributors and developers.
• NPM Workspaces and Publishing: Details on how the project's packages are managed and published.
• Stateless provider migration: Upgrade guide for the new provider runtime model.

Tools Documentation

• Tools Overview: Overview of the available tools.
• File System Tools: Documentation for the read_file and write_file tools.
• Multi-File Read Tool: Documentation for the read_many_files tool.
• Shell Tool: Documentation for the run_shell_command tool.
• MCP Server: Model Context Protocol server integration.
• Web Fetch Tool: Documentation for the web_fetch tool.
• Web Search Tool: Documentation for the google_web_search tool.
• Memory Tool: Documentation for the save_memory tool.

Additional Resources

• Release notes: Stateless Provider: Summary of new runtime features, breaking changes, and validation steps.
• Terms of Service and Privacy Notice: Information on the terms of service and privacy notices.

We hope this documentation helps you make the most of the LLxprt Code!