Overview
Lisply-MCP is a Model Context Protocol middleware that bridges MCP-enabled AI agents with Lisp-based development and runtime environments through the Lisply protocol. It can automatically start and manage a Lisply-compatible Docker backend, translate Lisp evaluation requests between the MCP interface and the Lisply API, and handle errors, Lisp debugger interaction, and logging. The wrapper exposes HTTP endpoints for Lisp evaluation, general HTTP requests, pings, and additional tooling; it supports both HTTP mode (structured responses with Result, Stdout, and Error) and a stdio mode that exposes the backend REPL stream for interactive debugging. Each server instance runs independently so you can operate multiple Lisply-backed Lisp environments in parallel without tool-name conflicts; tools are automatically prefixed with the server name. The default minimal setup uses a Gendl-based Common Lisp backend inside Docker, with optional Emacs SWANK integration and a range of advanced configuration options including mounting host directories, selecting Docker images and Lisp implementations, and configuring environment variables. Overall, it enables AI agents to generate, evaluate, load, test, and debug Lisp code in a flexible, containerized Lisp environment.
Features
Docker-based Lisply backend management
Automates starting and managing a Lisply-compliant Docker container for the Lisp backend.
MCP-Lisply request translation
Translates Lisp evaluation requests between MCP protocol and the backend Lisply API, handling errors and responses.
Dual communication modes: HTTP and STDIO
Supports HTTP mode with structured responses and STDIO mode for raw REPL interactions, including debugger support.
Per-server isolation and tool prefixing
Each MCP server runs independently; tools are prefixed with the server name to avoid conflicts.
SWANK Emacs integration
Includes a Lisply SWANK server bridge for Emacs connections.
Extensible configuration and image management
Offers docker image selection, Lisp implementation choice, mounting host directories, and environment variables for flexible setups.
Tool suite: Lisp evaluation, HTTP requests, and ping
Provides built-in MCP tools such as lisp_eval, http_request, and ping_lisp for interactive exploration.
Who Is This For?
- AI practitioners:Experiment with Lisp-enabled AI workflows; connect Claude Desktop to Lisp backends via MCP Lisply.
- Lisp practitioners:Leverage MCP Lisply to integrate Lisp environments with AI agents and tools.
- Neuro-Symbolic Programming researchers:Explore neuro-symbolic workflows by enabling AI agents to reason in Lisp.
- Engineers and CAD designers:Apply Lisp-based backends to automate CAD tasks with AI guidance.
- Tinkerers and hobbyists:Tinkerers and hobbyists exploring how Lisp backends can be controlled by AI agents.
- Claude Desktop users:Configure and test Lisply-MCP with Claude Desktop to drive Lisp environments.
