MCPLow riskUnclaimed

Vox MCP

Multi-model AI gateway — 10 providers including Cloudflare and Vercel, with conversation memory.

linxulelinxule/vox-mcp★ 1Updated Sep 26, 2026

server.json

{
  "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
  "name": "io.github.linxule/vox-mcp",
  "description": "Multi-model AI gateway — 10 providers including Cloudflare and Vercel, with conversation memory.",
  "title": "Vox MCP",
  "repository": {
    "url": "https://github.com/linxule/vox-mcp",
    "source": "github"
  },
  "version": "0.8.0",
  "packages": [
    {
      "registryType": "pypi",
      "identifier": "vox-mcp",
      "version": "0.8.0",
      "transport": {
        "type": "stdio"
      },
      "environmentVariables": [
        {
          "description": "Google Gemini API key",
          "isSecret": true,
          "name": "GEMINI_API_KEY"
        },
        {
          "description": "OpenAI API key",
          "isSecret": true,
          "name": "OPENAI_API_KEY"
        },
        {
          "description": "Anthropic API key",
          "isSecret": true,
          "name": "ANTHROPIC_API_KEY"
        },
        {
          "description": "xAI (Grok) API key",
          "isSecret": true,
          "name": "XAI_API_KEY"
        },
        {
          "description": "DeepSeek API key",
          "isSecret": true,
          "name": "DEEPSEEK_API_KEY"
        },
        {
          "description": "Moonshot (Kimi) API key",
          "isSecret": true,
          "name": "MOONSHOT_API_KEY"
        },
        {
          "description": "OpenRouter API key",
          "isSecret": true,
          "name": "OPENROUTER_API_KEY"
        },
        {
          "description": "Custom OpenAI-compatible endpoint URL (Ollama, vLLM, etc.)",
          "name": "CUSTOM_API_URL"
        },
        {
          "description": "Cloudflare API token with Workers AI Read permission",
          "isSecret": true,
          "name": "CLOUDFLARE_API_TOKEN"
        },
        {
          "description": "Cloudflare account ID (required with CLOUDFLARE_API_TOKEN)",
          "name": "CLOUDFLARE_ACCOUNT_ID"
        },
        {
          "description": "Cloudflare gateway ID (defaults to default)",
          "name": "CLOUDFLARE_GATEWAY_ID"
        },
        {
          "description": "Comma-separated Cloudflare upstream model IDs to list",
          "name": "CLOUDFLARE_MODELS"
        },
        {
          "description": "Vercel AI Gateway API key (AI_GATEWAY_API_KEY also accepted)",
          "isSecret": true,
          "name": "VERCEL_AI_GATEWAY_API_KEY"
        },
        {
          "description": "Alternative Vercel AI Gateway API key; VERCEL_AI_GATEWAY_API_KEY takes precedence",
          "isSecret": true,
          "name": "AI_GATEWAY_API_KEY"
        },
        {
          "description": "Comma-separated Vercel upstream model IDs to list",
          "name": "VERCEL_MODELS"
        },
        {
          "description": "Optional comma-separated Cloudflare model allowlist",
          "name": "CLOUDFLARE_ALLOWED_MODELS"
        },
        {
          "description": "Optional comma-separated Vercel model allowlist",
          "name": "VERCEL_ALLOWED_MODELS"
        },
        {
          "description": "Default model ID or alias; overrides saved settings. auto asks the agent to choose.",
          "name": "DEFAULT_MODEL"
        },
        {
          "description": "Optional path to non-secret model settings (default ~/.vox/config.json)",
          "name": "VOX_CONFIG_PATH"
        }
      ]
    }
  ]
}

Permissions

DeclaredDetected
Runs code—python
Installs—pypi:vox-mcp@0.8.0
Runs install scripts—None
NetworkNoneNone
Needs credentialsAI_GATEWAY_API_KEYANTHROPIC_API_KEYCLOUDFLARE_API_TOKENDEEPSEEK_API_KEYGEMINI_API_KEYMOONSHOT_API_KEYOPENAI_API_KEYOPENROUTER_API_KEYVERCEL_AI_GATEWAY_API_KEYXAI_API_KEYAI_GATEWAY_API_KEYANTHROPIC_API_KEYCLOUDFLARE_API_TOKENDEEPSEEK_API_KEYGEMINI_API_KEYMOONSHOT_API_KEYOPENAI_API_KEYOPENROUTER_API_KEYVERCEL_AI_GATEWAY_API_KEYXAI_API_KEY
Outside the workspace—None
Agent tools—None

Checks

Low risk · Nothing worth a warning was found.

Not reviewed by a person · Checked by rules; the model review is not switched on yet.

Versions

  1. #10.8.0latestOct 7, 2026