OGhidra integrates Large Language Models (LLMs) with Ghidra, a reverse engineering platform developed by the NSA. This tool enables users to conduct binary analysis using natural language commands, making it easier to interact with Ghidra’s features. Users can ask questions about functions and strings, automate tasks like renaming functions, and generate reports. The setup requires three main components: Ghidra, the OGhidraMCP plugin, and Ollama, a local LLM runtime. Python 3.12 or higher is needed, along with specific hardware recommendations, including at least 8GB of RAM and 50GB of free storage. The OGhidraMCP plugin enhances standard GhidraMCP by allowing multiple instances of Ghidra to run concurrently, facilitating parallel analysis. It includes features for raw byte reading and algorithmic scanning of data structures. Installation involves downloading the plugin, adding it through Ghidra’s extension manager, and configuring the server settings. Users can choose between two installation options: the recommended OGhidraMCP or the original GhidraMCP. Ollama powers the AI capabilities of OGhidra, supporting models like gpt-oss and gemma3 for effective binary analysis. Users need to pull these models and verify their installation. The OGhidra setup also provides options for using a modern Python package manager called UV or the traditional pip method to manage dependencies and run the application. This combination of tools simplifies reverse engineering tasks, making it accessible for both malware analysis and vulnerability research.