bethington/ghidra-mcp connects Ghidra to AI agents via a Model Context Protocol server, offering 256 tools for automating reverse engineering tasks, enforcing naming conventions, and executing batch operations.
Project overview
The project addresses the limited toolsets of most Ghidra MCP implementations by providing a large suite of tools along with automated convention enforcement within the tool layer.
Project type
MCP · Infrastructure
Use cases
Coding & Development
Deployment
Refer to project documentation
License
Apache-2.0
Best for
Developers using AI agents for reverse engineering tasks who need to enforce naming and type conventions within Ghidra.
Key capabilities
Enforces naming and type conventions in the tool layer with auto-fix, warn, or reject behaviors to ensure consistent output across AI agents.
Supports loading only default tool groups on connection to reduce model context overhead, with dynamic discovery and loading of other tools.
Limitations and risks
Without strict program routing, MCP calls might silently write to the wrong binary if multiple programs are open simultaneously.
Getting started
Setup difficulty is medium due to prerequisites such as Ghidra, Java 21, Maven, and Python. The first success path involves cloning the repository, running 'python -m tools.setup ensure-prereqs', and running 'python -m tools.setup build'.
Users can interact with the server via command-line interface or API.
Evidence and sources
README: A production-ready Model Context Protocol (MCP) server that bridges Ghidra's powerful reverse engineering capabilities with modern AI tools and automation frameworks. **256 MCP to…
README: Most Ghidra MCP implementations give you a handful of read-only tools and call it a day. This project is different
README: v5.0 moves conventions from "things to remember" into the tool layer, where they can actually be enforced.
Release: GhidraMCP 5.14.2
README: Function Analysis — Decompilation, call graphs, cross-references, completeness scoring