mcp-server-code-execution-mode

# MCP Code Execution Server: Zero-Context Discovery for 100+ MCP Tools [](https://mseep.ai/app/elusznik-mcp-server-code-execution-mode) **Stop paying 30,000 tokens per query.** This bridge implements Anthropic's discovery pattern with rootless security—reducing MCP context from 30K to 200 tokens while proxying any stdio server. [](https://www.anthropic.com/engineering/code-execution-with-mcp) [](https://blog.cloudflare.com/code-mode/) [](https://www.docker.com/blog/dynamic-mcps-stop-hardcoding-your-agents-world/) [](https://machinelearning.apple.com/research/codeact) [](https://modelcontextprotocol.io/) [](https://mseep.ai/app/4a84c349-1795-41fc-a299-83d4a29feee8) ## Overview This bridge implements the **"Code Execution with MCP"** pattern, a convergence of ideas from industry leaders: - **Apple's [CodeAct](https://machinelearning.apple.com/research/codeact)**: "Your LLM Agent Acts Better when Generating Code." - **Anthropic's [Code execution with MCP](https://www.anthropic.com/engineering/code-execution-with-mcp)**: "Building more efficient agents." - **Cloudflare's [Code Mode](https://blog.cloudflare.com/code-mode/)**: "LLMs are better at writing code to call MCP, than at calling MCP directly." - **Docker's [Dynamic MCPs](https://www.docker.com/blog/dynamic-mcps-stop-hardcoding-your-agents-world/)**: "Stop Hardcoding Your Agents’ World." - **[Terminal Bench](https://www.tbench.ai)'s [Terminus](https://www.tbench.ai/terminus)**: "A realistic terminal environment for evaluating LLM agents." Instead of exposing hundreds of individual tools to the LLM (which consumes massive context and confuses the model), this bridge exposes **one** tool: `run_python`. The LLM writes Python code to discover, call, and compose other tools. ### Why This vs. JS "Code Mode"? While there are JavaScript-based alternatives (like [`universal-tool-calling-protocol/code-mode`](https://github.com/universal-tool-calling-protocol/code-mode)), this project is built for **Data Science** and **Security**: | Feature | This Project (Python) | JS Code Mode (Node.js) | | :--- | :--- | :--- | | **Native Language** | **Python** (The language of AI/ML) | TypeScript/JavaScript | | **Data Science** | **Native** (`pandas`, `numpy`, `scikit-learn`) | Impossible / Hacky | | **Isolation** | **Hard** (Podman/Docker Containers) | Soft (Node.js VM) | | **Security** | **Enterprise** (Rootless, No Net, Read-Only) | Process-level | | **Philosophy** | **Infrastructure** (Standalone Bridge) | Library (Embeddable) | **Choose this if:** You want your agent to analyze data, generate charts, use scientific libraries, or if you require strict container-based isolation for running untrusted code. ## What This Solves (That Others Don't) ### The Pain: MCP Token Bankruptcy Connect Claude to 11 MCP servers with ~100 tools = **30,000 tokens** of tool schemas loaded into every prompt. That's **$0.09 per query** before you ask a single question. Scale to 50 servers and your context window *breaks*. ### Why Existing "Solutions" Fail - **Docker MCP Gateway**: Manages containers beautifully, but still streams **all tool schemas** into Claude's context. No token optimization. - **Cloudflare Code Mode**: V8 isolates are fast, but you **can't proxy your existing MCP servers** (Serena, Wolfram, custom tools). Platform lock-in. - **Academic Papers**: Describe Anthropic's discovery pattern, but provide **no hardened implementation**. - **Proofs of Concept**: Skip security (no rootless), skip persistence (cold starts), skip proxying edge cases. ### The Fix: Discovery-First Architecture - **Constant 200-token overhead** regardless of server count - **Proxy any stdio MCP server** into rootless containers - **Fuzzy search across servers** without preloading schemas - **Production-hardened** with capability dropping and security isolation ### Architecture: How It Differs ```text Traditional MCP (Context-Bound) ┌─────────────────────────────┐ │ LLM Context (30K tokens) │ │ - serverA.tool1: {...} │ │ - serverA.tool2: {...} │ │ - serverB.tool1: {...} │ │ - … (dozens more) │ └─────────────────────────────┘ ↓ LLM picks tool ↓ Tool executes This Bridge (Discovery-First) ┌─────────────────────────────┐ │ LLM Context (≈200 tokens) │ │ “Use discovered_servers(), │ │ query_tool_docs(), │ │ search_tool_docs()” │ └─────────────────────────────┘