PaulDHaes/Ghostwriter-mcp

Ghostwriter-mcp

SpecterOps Ghostwriter MCP server

Installation

Install the pip packages by going into the directory

pip install -r requirements.txt

If this does not work use a virtual env

mkdir ghostwriter-python
cd ghostwriter-python
python3 -m venv ghostwriter-env
source ghostwriter-env/bin/activate
cd ..

To run this can be done and imported to a mcp server using stdio by simply doing this in the mcp config

{
  "Ghostwriter-mcp": {
    "command": "/Users/username/Ghostwriter-mcp/ghostwriter-python/ghostwriter-env/bin/python3",
    "args": [
      "/Users/username/Ghostwriter-mcp/main.py"
    ],
    "env": {},
    "working_directory": null
  }
}

If your LLM wants to use it over streamable http try using mcpo

uvx mcpo --port 8000 -- python3 main.py

Try connecting to the ip you execute the script from+new port

http://127.0.0.1:8000

What can it do

It should be able to create a new client,project and report. It then can also attach findings to a report + update the findings for now only the evidence and affected entities part. It can generate codename for clients and projects. It should also be ables to search for clienst,project,reports and findings for avoiding duplication problems.