Pharo NeoConsole

Pharo NeoConsole is an MCP server that connects AI assistants to a local Pharo Smalltalk environment using a NeoConsole server backend. Designed for Smalltalk developers, software engineers, and researchers working with Pharo images, this integration allows language models to directly interact with dynamic Smalltalk runtime sessions. By exposing specialized tools over the Model Context Protocol, the server enables users to evaluate complex Smalltalk expressions, run one-off scripts using Pharo command-line options, and query live system health metrics like memory usage and runtime status. Developers can also perform reflective code analysis directly from their AI-assisted development environments. The server can fetch class comments, retrieve class definitions, list method selectors across classes, and read exact method source code from the active Pharo image. This creates a bridge between modern AI developer tools and the interactive Smalltalk live programming environment, removing the friction of manually switching between code editors and Pharo workspaces during development and debugging workflows.

Category: Cloud & Infrastructure

Tags: code-execution, pharo, repl, smalltalk

Visit Pharo NeoConsole

How to install and configure Pharo NeoConsole

1. Ensure Python 3.10 or later, the uv package manager, and Pharo with NeoConsole are installed on your system. 2. Set the PHARO_DIR environment variable to your Pharo directory (defaults to ~/pharo) and ensure NeoConsole.image is present. 3. Clone the repository and install dependencies by running uv sync --dev. 4. Add the server configuration to your Cursor mcp.json configuration file: json { "mcpServers": { "pharo-nc-mcp-server": { "command": "uv", "args": [ "--directory", "/your-path/to/pharo-nc-mcp-server", "run", "pharo-nc-mcp-server" ] } } } 5. Alternatively, start the MCP server manually from the terminal using uv run pharo-nc-mcp-server.

What you can do with Pharo NeoConsole

• Evaluate Smalltalk expressions inside a running Pharo image through NeoConsole to test dynamic behavior and algorithms. - Retrieve system runtime status and free memory metrics to monitor resource utilization inside the active Pharo environment. - Inspect Pharo class definitions and comments to understand class hierarchies and internal object models directly in your AI assistant. - Search method lists and fetch source code for specific class selectors to analyze implementations and debug existing Smalltalk logic.

Key facts

• Open Source
• https://github.com/svenvc/NeoConsole
• Cloud & Infrastructure, Developer Tools & Code Intelligence
• code-execution, pharo, repl, smalltalk

Part of MCP Servers

What can Pharo NeoConsole do?

Pharo NeoConsole enables AI assistants to interact directly with a running Pharo Smalltalk image. It provides tools to evaluate Smalltalk expressions via NeoConsole or simple evaluation commands, inspect system metrics such as memory and status, and query code structure. Users can retrieve class comments, class definitions, method lists, and method source code without leaving their AI-driven development environment.

How do I install Pharo NeoConsole?

You can install Pharo NeoConsole by cloning its repository and running uv sync to install Python dependencies. You must have Python 3.10 or later, the uv package manager, and a Pharo installation containing NeoConsole. Point the PHARO_DIR environment variable to your Pharo directory and register the server command in your client configuration file.

Which MCP clients work with Pharo NeoConsole?

Pharo NeoConsole works with any client that supports the Model Context Protocol, including Cursor and Claude Desktop. Configuration involves defining the server command using uv and passing the directory path to the cloned repository in your client configuration file, such as mcp.json.

Is Pharo NeoConsole open source?

Yes, Pharo NeoConsole is an open source project. Developers can freely review the code, inspect its tool implementations, contribute fixes, run automated test suites, and customize its tools according to their own Smalltalk development requirements and workflow needs.