xa — Fastest Coding-Agent CLI

Execute Anything. Pure Rust. BYOK. Zero bloat.

xa is a blazing-fast, pure-Rust coding agent CLI. It combines the power of LLM-driven automation with the raw performance of Rust — delivering the fastest possible response times, minimal token consumption, and complete flexibility in provider selection.

Why xa?

• 🚀 Pure Rust — Built entirely in Rust for extreme speed and zero dependency overhead. No Python, no Node.js, just native performance.
• ⚡ Fastest Speed — Minimal latency from input to output. The lightweight architecture and efficient streaming ensure you get results faster than any web-based alternative.
• 🔑 BYOK (Bring Your Own Key) — Full support for your own API keys. No vendor lock-in, no forced accounts.
• 🌐 Any Custom Endpoint — Works with any OpenAI-compatible API: OpenAI, OpenRouter, vLLM, llama.cpp, Ollama, local gateways, or self-hosted models. Just plug in your endpoint.
• 💰 Minimal Token Usage — Built-in RTK (Real-Time Knowledge) token minimization post-processing module strips unnecessary tokens from responses, reducing your API costs significantly.
• 🖥️ Interactive TUI — Codex-style terminal UI with streaming markdown rendering, thinking indicators, slash commands, session management, and smart input handling.
• 📋 Command-Line Power — One-liner text processing: translate, polish, summarize, rewrite — all from your terminal.
• 🔍 Fuzzy Command Matching — Type partial commands and let xa figure out your intent.
• 📎 Clipboard Integration — Results are automatically copied to your system clipboard.

Installation

# Clone the repository
git clone https://github.com/jinfagang/xa.git
cd xa

# Build (requires Rust toolchain)
cargo build --release

# Install globally (optional)
cargo install --path .

# Binary available at target/release/xa

Quick Start

1. Login / Configure Provider

# Interactive login wizard (supports any OpenAI-compatible endpoint)
xa login

During setup, you'll be prompted for:

• Endpoint URL — e.g. https://api.openai.com/v1, https://openrouter.ai/api/v1, or any custom endpoint
• API Key — your own key (BYOK)
• Model — choose from available models or specify a custom one

2. Chat (Interactive TUI)

# Launch the codex-like interactive coding agent
xa

Inside the TUI:

Shortcut	Action
Enter	Send message
Shift+Enter	Newline
Ctrl+C	Stop streaming
/login [name]	Add/update a provider
/models [name]	Switch provider or model
/clear	Clear conversation
/save [title]	Save session
/sessions	List saved sessions
/help	Show all commands

3. Session Management

# Save current session
xa --session save "My conversation"

# List sessions
xa --session ls

# Resume a session
xa --session <id>

Architecture

┌──────────────────────────────────────────────┐
│                 xa CLI                        │
│                                              │
│  ┌───��──────┐  ┌──────────┐  ┌────────────┐ │
│  │  TUI     │  │  LLM     │  │  Token     │ │
│  │(ratatui) │→ │(stream)  │→ │(RTK filter)│ │
│  └──────────┘  └──────────┘  └────────────┘ │
│       ↓              ↓              ↓        │
│  HistoryCells   Any Provider   Minimal Tokens│
└──────────────────────────────────────────────┘

• TUI Layer: Built on ratatui + crossterm with an Elm-style event loop. Virtual scrolling, markdown rendering, thinking indicators, and slash-command overlays.
• LLM Layer: Abstraction over any OpenAI-compatible chat completions API. Streaming and non-streaming modes.
• Token Module: Built-in RTK (Real-Time Knowledge) token minimization post-processing — strips verbose output, reduces token waste.

Configuration

File	Purpose
~/.config/xa/config.toml	Default API settings
~/.config/xa/providers.toml	Multi-provider management
~/.config/xa/prompts.toml	Custom prompt templates
~/.config/xa/sessions/	Saved conversation sessions

Supported Providers

xa works with any OpenAI-compatible endpoint:

• OpenAI — https://api.openai.com/v1
• OpenRouter — https://openrouter.ai/api/v1
• Ollama — http://localhost:11434/v1
• vLLM — custom deployment
• llama.cpp — server mode
• Any custom endpoint — just configure it

No hardcoded providers. No restrictions. Your model, your rules.

Performance

Metric	xa	Alternatives
Startup time	~5ms	~200ms+
Memory footprint	~10MB	~100MB+
Language	Pure Rust	Python/Node
Dependencies	Minimal	Heavy

License

MIT