@abdoknbgit/claudex

Claudex

The multi-provider AI coding CLI. Use Claude Code's full agentic tool loop, MCP servers, hooks, skills, and interactive TUI with every major LLM provider.

Claudex is not a proxy or wrapper. It is a complete reimplementation of the Claude Code runtime with a provider-agnostic core: every provider goes through native adapters that translate the Anthropic tool-use protocol into each provider's native API format, with full streaming, rate-limit handling, and retry logic built in.

Supported Providers: Anthropic, OpenAI, Google Gemini, Groq, DeepSeek, Ollama, NVIDIA NIM, OpenRouter

Quick Install

npm install -g @abdoknbgit/claudex

From source

git clone https://github.com/AbdoKnbGit/claudex.git
cd claudex
npm install
npm run build
npm link

Requirements

• Node.js >= 20.0.0
• Git (optional on Windows — claudex auto-falls-back to PowerShell if git-bash is absent)
• Bun (optional — npm run build uses esbuild; npm run build:bun uses Bun)

Launch

# Interactive mode (full TUI)
claudex

# One-shot print mode
claudex -p "explain this codebase"

# Continue last conversation
claudex -c

Commands

/provider - Switch between LLM providers

Opens an interactive picker to switch between all 8 supported providers. Your selection is saved and persists across sessions. Each provider uses its native API format for best performance.

/models - Browse and pick any model

Opens a live model browser that fetches the full model catalog from the selected provider's API. Search, filter, and set any model as your active model.

• /models — open interactive provider + model picker
• /models <query> — search the active provider's models
• /models <provider>:<query> — search a specific provider (e.g. /models groq:llama)
• /model <model-id> — set a specific model directly (e.g. /model deepseek-reasoner)

Example model counts per provider:

Provider	Models Available
OpenAI	80+
Google Gemini	30+
OpenRouter	300+
Groq	15+
DeepSeek	5+
NVIDIA NIM	100+
Ollama	varies (local)

/thinking - Toggle thinking/reasoning mode

Controls whether the model reasons step-by-step before answering. Works safely across all providers:

• /thinking — toggle on/off
• /thinking on — enable thinking
• /thinking off — disable thinking
• /thinking status — show current state and model support

Decision matrix (zero crashes guaranteed):

State	Model Support	Behavior
thinking=on	model supports thinking	Thinking enabled
thinking=on	model lacks thinking	Request sent normally (param silently omitted)
thinking=off	any model	No thinking param sent

Models with thinking support: Claude 4+ (Anthropic), DeepSeek Reasoner, Kimi K2 Thinking (NIM), DeepSeek R1 distill models (Groq)

Features

Multi-Provider Engine

• 8 providers with native adapters (not a proxy)
• OAuth + API key auth for all providers
• Persistent provider selection across sessions
• Automatic tool schema sanitization per provider (Gemini, OpenAI, Groq)
• Aggressive payload optimization for rate-limited providers (Groq)
• Human-friendly error messages for billing, quota, and auth errors

Full Agent Tooling

• File editing — Read, Edit, Write, Glob, Grep with permission controls
• Bash execution — sandboxed shell commands with timeout and abort
• MCP servers — full Model Context Protocol support for external tools
• Hooks — PreToolUse, PostToolUse, UserPromptSubmit, Stop, Notification
• Skills — /commit, /review-pr, /simplify, /loop, and more
• Task management — structured task tracking within sessions
• Web tools — WebSearch, WebFetch for live data retrieval

Developer Experience

• Interactive TUI — Ink-based React terminal UI with streaming output
• Thinking mode — toggle reasoning with /thinking across all providers
• Vim mode — modal editing in the prompt
• Keyboard shortcuts — configurable keybindings
• Debug mode — claudex --debug for full request/response tracing
• VS Code extension — companion extension with Control Center, provider switching, project-aware launch

Resilience

• Retry with backoff — up to 10 retries with exponential backoff + jitter for 429/5xx errors
• Retry-After respect — honors provider rate-limit headers
• Context overflow recovery — auto-reduces max_tokens when hitting context limits
• Streaming fallback — NIM and Ollama fall back to non-streaming when needed
• Cross-platform, shell-agnostic — npm install -g works on Windows, macOS, Linux, WSL with no extra setup. On Windows, claudex uses git-bash if present and auto-falls-back to PowerShell otherwise. No environment variables required — provider credentials are stored via a guided first-run wizard.

VS Code Extension

The claudex-vscode/ directory contains a companion VS Code extension:

• Control Center — webview panel showing provider state, project info, and launch actions
• Provider switching — change providers from the VS Code command palette
• Project-aware launch — starts claudex in the right directory with the right context
• Status bar — shows active provider and session state

Install from the extension directory:

cd claudex-vscode
npx @vscode/vsce package --no-dependencies
code --install-extension claudex-vscode-*.vsix

License

MIT