Featured

Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger
Crawl and scrape any site into clean data, 10% off logoCrawl and scrape any site into clean data, 10% off

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits, and new users get 10% off their first purchase.

Try Firecrawl free
Your own AI agent, running 24/7 with QwikClaw logoYour own AI agent, running 24/7 with QwikClaw

QwikClaw sets up and runs an always-on OpenClaw agent for you. One click, no config files, no server setup.

Deploy now
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free
SetupClaw: done-for-you OpenClaw for founders & exec teams logoSetupClaw: done-for-you OpenClaw for founders & exec teams

White-glove OpenClaw for founders and exec teams (4–50+ employees): we install, harden, integrate your tools, and maintain it — secured from day one.

Get it set up for you
SEO data APIs for your agent, $1 free credit logoSEO data APIs for your agent, $1 free credit

DataForSEO gives your agent live access to SERP results, keyword data, backlinks, and on-page SEO data through one API. New accounts get a $1 credit, good for up to 20,000 keyword or backlink lookups.

Try DataForSEO free
Reach 47,000+ AI builders

A flat monthly placement in front of developers actively installing AI tools. No lock-in, cancel anytime.

Advertise here

Summary

Ghidra reverse engineering integration - analyze binaries, decompile functions, find strings, trace cross-references

Install to Claude Code

/plugin install ghidra@ghidra-tools

Run in Claude Code. Add the marketplace first with /plugin marketplace add Und3rf10w/ai-ghidra-tools if you haven't already.

README.md

Ghidra Plugin for Claude Code

A comprehensive reverse engineering integration that brings Ghidra's powerful binary analysis capabilities directly into Claude Code through the Model Context Protocol (MCP).

Overview

This plugin enables Claude to analyze executables, decompile functions, search for patterns, trace cross-references, and even emulate code—all through natural language conversations. It bridges Ghidra's headless analyzer with Claude Code, making reverse engineering tasks more accessible and efficient.

Key Capabilities

  • Binary Analysis: Import and analyze executables, libraries, and firmware
  • Decompilation: Convert assembly to readable C pseudocode
  • Disassembly: View raw assembly instructions with full context
  • Pattern Searching: Find byte sequences, strings, and signatures
  • Cross-References: Trace data and code references throughout the binary
  • Control Flow: Analyze basic blocks and call graphs
  • Emulation: Execute code with Ghidra's P-code emulator
  • Annotation: Add comments and rename symbols to document findings

Requirements

  • Ghidra 11.0+ (tested with 12.0.2)
  • Python 3.9+
  • Java 17+ (for Ghidra)
  • Claude Code CLI

Installation

1. Install Ghidra

Download and extract Ghidra from the official releases.

# Example: Extract to ~/ghidra
unzip ghidra_12.0.2_PUBLIC.zip -d ~/

2. Set Environment Variable

# Add to your shell profile (~/.bashrc, ~/.zshrc, etc.)
export GHIDRA_INSTALL=~/ghidra_12.0.2_PUBLIC

3. Install the Plugin

Clone or copy the plugin to your Claude Code plugins directory:

# Clone the repository
git clone https://github.com/your-org/ghidra-claude-plugin.git

# Or copy to plugins directory
cp -r ghidra ~/.claude/plugins/

4. Install Python Dependencies

cd ~/.claude/plugins/ghidra
python3 -m venv .venv
source .venv/bin/activate
pip install mcp httpx

5. Verify Installation

claude -p "analyze /bin/ls and list its functions"

Configuration

Environment Variables

| Variable | Description | Default | |----------|-------------|---------| | GHIDRA_INSTALL | Path to Ghidra installation | ~/ghidra_*_PUBLIC | | GHIDRA_WORKSPACE | Directory for Ghidra projects | ~/.claude/ghidra-workspace |

MCP Server Configuration

The plugin's .mcp.json configures the MCP server:

{
  "mcpServers": {
    "ghidra-mcp": {
      "command": "python3",
      "args": ["${CLAUDE_PLUGIN_ROOT}/servers/ghidra-mcp/server.py"],
      "env": {
        "GHIDRA_INSTALL": "${GHIDRA_INSTALL}",
        "GHIDRA_WORKSPACE": "${HOME}/.claude/ghidra-workspace"
      }
    }
  }
}

Available Tools

Analysis & Navigation

| Tool | Description | |------|-------------| | analyze_binary | Import and analyze a binary file. Returns metadata including architecture, entry point, and function count. Must be called first. | | list_functions | List functions with pagination and filtering. Returns names, addresses, sizes, and signatures. | | get_symbols | Get imported and exported symbols. Shows external dependencies and exports. | | list_classes | List C++/Objective-C classes, vtables, and methods. |

Decompilation & Disassembly

| Tool | Description | |------|-------------| | decompile_function | Decompile a function to C pseudocode. Returns code, signature, and local variables. | | get_disassembly | Get raw assembly for a function or address range. Includes bytes, mnemonics, and operands. | | get_basic_blocks | Get control flow graph with basic blocks, edges, and instructions. |

Search & References

| Tool | Description | |------|-------------| | search_strings | Find strings in the binary with minimum length and pattern filtering. | | search_bytes | Search for byte patterns with wildcard support (e.g., 48 ?? 05). | | get_xrefs | Get cross-references to/from an address. Trace callers, callees, and data refs. | | get_call_graph | Get caller/callee tree for understanding function relationships. |

Memory & Data

| Tool | Description | |------|-------------| | get_memory_map | Get memory sections with permissions and addresses. | | get_data_at_address | Read and interpret data at an address (bytes, integers, pointers, strings). |

Modification & Annotation

| Tool | Description | |------|-------------| | rename_symbol | Rename a function or symbol. Changes persist in the Ghidra project. | | add_comment | Add comments at addresses (EOL, pre, post, plate, repeatable). | | set_function_signature | Update function return type and parameters. | | patch_bytes | Modify bytes at an address for patching or experimentation. |

Advanced

| Tool | Description | |------|-------------| | emulate_function | Execute code using Ghidra's P-code emulator with custom register/memory inputs. |

Usage Examples

Basic Analysis

User: Analyze /bin/ls and show me its main function

Claude: [Calls analyze_binary, then decompile_function]
        Here's the decompiled main function...

Finding Vulnerabilities

User: Search for strcpy calls in this binary

Claude: [Calls search_strings to find "strcpy", then get_xrefs]
        Found 3 calls to strcpy at addresses...

Understanding Control Flow

User: Show me the control flow graph for the authentication function

Claude: [Calls get_basic_blocks]
        The function has 12 basic blocks with the following structure...

Pattern Hunting

User: Find all instances of the byte pattern "48 89 e5" (mov rbp, rsp)

Claude: [Calls search_bytes]
        Found 47 matches, primarily at function prologues...

Crypto Analysis with Emulation

User: Emulate the encrypt function with input "test"

Claude: [Calls emulate_function with memory setup]
        After 156 steps, RAX contains 0x7a3b2c1d...

Project Structure

ghidra/
├── .claude-plugin/
│   └── plugin.json           # Plugin metadata
├── .mcp.json                  # MCP server configuration
├── servers/
│   └── ghidra-mcp/
│       ├── server.py         # MCP server implementation
│       └── ghidra_bridge.py  # Ghidra headless wrapper
├── ghidra_scripts/           # Jython scripts for Ghidra
│   ├── analyze_binary.py
│   ├── decompile_function.py
│   ├── get_disassembly.py
│   ├── search_bytes.py
│   ├── emulate_function.py
│   └── ... (18 scripts total)
├── commands/                 # Slash commands
│   ├── analyze.md
│   └── decompile.md
├── agents/                   # Specialized agents
│   └── reverse-engineer.md
├── skills/                   # Domain knowledge
│   └── ghidra-scripting/
├── tests/
│   └── test_ghidra_mcp.py   # Pytest test suite
└── README.md

Development

Running Tests

cd ~/.claude/plugins/ghidra
source .venv/bin/activate
pip install pytest

# Run read-only tests
pytest -m "not write" -v

# Run all tests (includes write operations)
pytest -v

Adding New Tools

1. Create a Jython script in ghidra_scripts/:

# @category Claude.MCP
# @runtime Jython

import json

def output_json(data):
    print("===JSON_START===")
    print(json.dumps(data))
    print("===JSON_END===")

def run():
    # Your implementation
    output_json({"status": "success", "data": result})

run()

2. Add wrapper method in ghidra_bridge.py 3. Add Tool definition in server.py 4. Add tests in test_ghidra_mcp.py

Debugging

Enable verbose output by checking:

  • Ghidra log: ~/Library/ghidra/ghidra_*/application.log
  • Script stdout: Captured in bridge's _parse_json_output

Supported Architectures

The plugin supports all architectures that Ghidra supports, including:

  • x86 / x86-64 (Intel/AMD)
  • ARM / ARM64 (AArch64)
  • MIPS / MIPS64
  • PowerPC
  • SPARC
  • RISC-V
  • 68000 (Motorola)
  • 8051
  • AVR
  • And 20+ more

Supported File Formats

  • Executables: ELF, PE/COFF, Mach-O
  • Libraries: .so, .dll, .dylib
  • Firmware: Raw binaries, Intel HEX, Motorola S-records
  • Archives: .a, .lib
  • Other: DEX (Android), Java class files, PDB symbols

Troubleshooting

"Ghidra analyzeHeadless not found"

Ensure GHIDRA_INSTALL points to your Ghidra installation:

export GHIDRA_INSTALL=/path/to/ghidra_12.0.2_PUBLIC

"No program loaded"

Call analyze_binary first before using other tools:

User: List functions in /bin/ls
Claude: I'll first analyze the binary, then list its functions.
        [Calls analyze_binary, then list_functions]

Analysis is slow

Large binaries may take several minutes. Use the timeout parameter:

analyze_binary("/path/to/large_binary", timeout=900)  # 15 minutes

Memory issues

Increase Ghidra's heap size by setting in your environment:

export GHIDRA_HEADLESS_MAXMEM=8G

Security Considerations

  • This plugin executes Ghidra's headless analyzer on binaries
  • Only analyze binaries from trusted sources
  • The plugin creates projects in ~/.claude/ghidra-workspace/
  • Patch operations modify the Ghidra project, not the original binary

Related plugins

Browse all →