ask llm for Claude Code

io.github.lykhoyda/ask-llm

Unified MCP server for Gemini, Codex, Claude, Grok, Ollama, and Antigravity

client:Claude Code transport:stdio runtime:npm

Install ask llm in Claude Code

run in your project directory

XAI_API_KEY

ASK_GROK_HARNESS

ASK_GROK_MODEL

ASK_GROK_REASONING_EFFORT

ASK_GROK_MAX_OUTPUT_TOKENS

ASK_GROK_TIMEOUT_MS

CURSOR_API_KEY

ASK_CURSOR_TIMEOUT_MS

ASK_CLAUDE_MODEL

ASK_CLAUDE_FALLBACK_MODEL

ASK_CLAUDE_TIMEOUT_MS

OLLAMA_HOST

ASK_ANTIGRAVITY_TIMEOUT_MS

ASK_ANTIGRAVITY_SANDBOX

GMCPT_TIMEOUT_MS

GMCPT_LOG_LEVEL

Held in this page only — never stored, logged, or sent anywhere but back to your screen.

claude mcp add ask-llm -e XAI_API_KEY=<XAI_API_KEY> -e ASK_GROK_HARNESS=<ASK_GROK_HARNESS> -e ASK_GROK_MODEL=<ASK_GROK_MODEL> -e ASK_GROK_REASONING_EFFORT=<ASK_GROK_REASONING_EFFORT> -e ASK_GROK_MAX_OUTPUT_TOKENS=<ASK_GROK_MAX_OUTPUT_TOKENS> -e ASK_GROK_TIMEOUT_MS=<ASK_GROK_TIMEOUT_MS> -e CURSOR_API_KEY=<CURSOR_API_KEY> -e ASK_CURSOR_TIMEOUT_MS=<ASK_CURSOR_TIMEOUT_MS> -e ASK_CLAUDE_MODEL=<ASK_CLAUDE_MODEL> -e ASK_CLAUDE_FALLBACK_MODEL=<ASK_CLAUDE_FALLBACK_MODEL> -e ASK_CLAUDE_TIMEOUT_MS=<ASK_CLAUDE_TIMEOUT_MS> -e OLLAMA_HOST=<OLLAMA_HOST> -e ASK_ANTIGRAVITY_TIMEOUT_MS=<ASK_ANTIGRAVITY_TIMEOUT_MS> -e ASK_ANTIGRAVITY_SANDBOX=<ASK_ANTIGRAVITY_SANDBOX> -e GMCPT_TIMEOUT_MS=<GMCPT_TIMEOUT_MS> -e GMCPT_LOG_LEVEL=<GMCPT_LOG_LEVEL> -- npx -y @ask-llm/mcp

Adds it for this project only. Append --scope user to make it available everywhere. Claude Code docs

ask llm in other clients

ask llm MCP server