MCP.so
ログイン
S

Secure LLM Proxy

@wootock

リポジトリ

Secure LLM Proxy について

OpenAI-compatible proxy with PII scrubbing, audit logging, and model routing. Drop-in replacement — set OPENAI_BASE_URL to point to this server. Supports GPT-4o, Claude, Gemini, and Llama via single endpoint.

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "secure-llm-proxy": {
      "url": "https://nexus.exchange-secure.com/mcp",
      "headers": {
        "Authorization": "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"
      }
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Secure LLM Proxy?

Secure LLM Proxy is an OpenAI-compatible proxy server that adds privacy controls, including PII scrubbing and audit logging, before forwarding requests to upstream LLM providers. It is designed for users who need a drop-in replacement for the OpenAI API with enhanced data protection.

How to use Secure LLM Proxy?

Set OPENAI_BASE_URL to https://nexus.exchange-secure.com/v1 and provide your API key. For MCP integration, configure the server with that URL and a Bearer token in your client’s settings. Use the chat_completion tool to send messages through the proxy.

Key features of Secure LLM Proxy

  • PII scrubbing removes emails, phone numbers, SSNs, and credit card numbers.
  • Audit logging records every request with a hash of the original content.
  • OpenAI-compatible endpoint — just change OPENAI_BASE_URL.
  • Model routing supports GPT-4o, Claude, Gemini, and Llama via a single endpoint.
  • Rate limit pooling spreads load across multiple upstream keys.

Use cases of Secure LLM Proxy

  • Send sensitive prompts to LLMs without exposing personal data.
  • Maintain an auditable log of all LLM interactions in regulated environments.
  • Route requests to multiple LLM providers from a single, privacy-preserving gateway.
  • Load-balance across several API keys to avoid rate limits.

FAQ from Secure LLM Proxy

What PII does the proxy scrub?

It strips emails, phone numbers, social security numbers, and credit card numbers before forwarding the request.

Are my API keys stored by the proxy?

No, keys are never stored. They are used only to forward the request after PII scrubbing.

Where do audit logs live and what do they contain?

Every request is logged with a hash of the original content. The full content is not stored; only the hash is retained for audit purposes.

Which models does the proxy support?

It supports GPT-4o, Claude, Gemini, and Llama through a single endpoint. Model routing is handled automatically based on the model parameter.

How do I authenticate with the MCP server?

Include your API key as a Bearer token in the Authorization header when configuring the MCP server in your client.

よくある質問

What PII does the proxy scrub?

It strips emails, phone numbers, social security numbers, and credit card numbers before forwarding the request.

Are my API keys stored by the proxy?

No, keys are never stored. They are used only to forward the request after PII scrubbing.

Where do audit logs live and what do they contain?

Every request is logged with a hash of the original content. The full content is not stored; only the hash is retained for audit purposes.

Which models does the proxy support?

It supports GPT-4o, Claude, Gemini, and Llama through a single endpoint. Model routing is handled automatically based on the `model` parameter.

How do I authenticate with the MCP server?

Include your API key as a Bearer token in the `Authorization` header when configuring the MCP server in your client.

コメント

基本情報

カテゴリ

AI とエージェント

トランスポート

stdio

作者

wootock

投稿者

Adam T

「AI とエージェント」の他のコンテンツ

MCP-NixOS - Because Your AI Assistant Shouldn't Hallucinate About Packages

utensils

MCP-NixOS - Model Context Protocol Server for NixOS resources

739

Open Multi-Agent Canvas

CopilotKit

The open-source multi-agent chat interface that lets you manage multiple agents in one dynamic conversation and add MCP servers for deep research

507

Mcp Agent

lastmile-ai

Build effective agents using Model Context Protocol and simple workflow patterns

8.4K

欢迎来到 智言平台

Shy2593666979

AgentChat 是一个基于 LLM 的智能体交流平台，内置默认 Agent 并支持用户自定义 Agent。通过多轮对话和任务协作，Agent 可以理解并协助完成复杂任务。项目集成 LangChain、Function Call、MCP 协议、RAG、Memory、HITL、Skill、Milvus 和 ElasticSearch 等技术，实现高效的知识检索与工具调用，使用 FastAPI 构建高性能后端服务。

797

Lumify Sports Intelligence

Lumify

Real-time sports schedules, live scores, odds, betting splits, and AI bet intelligence for agents via MCP.