Manticore is a prototyping tool for dynamic binary analysis, with support for symbolic execution, taint analysis, and binary instrumentation.
- Input Generation: Manticore automatically generates inputs that trigger unique code paths
- Crash Discovery: Manticore discovers inputs that crash programs via memory safety violations
- Execution Tracing: Manticore records an instruction-level trace of execution for each generated input
- Programmatic Interface: Manticore exposes programmatic access to its analysis engine via a Python API
Manticore supports binaries of the following formats, operating systems, and architectures. It has been primarily used on binaries compiled from C and C++. Examples of practical manticore usage are also on github.
- OS/Formats: Linux ELF
- Architectures: x86, x86_64, ARMv7
Manticore is supported on Linux, and requires Python 2.7. Ubuntu 16.04 is strongly recommended.
Install and try Manticore in a few shell commands (see an asciinema):
# Install system dependencies
sudo apt-get update && sudo apt-get install python-pip -y
python -m pip install -U pip
# Install manticore and its dependencies
sudo pip install manticore
# Download and build the examples
git clone https://github.com/trailofbits/manticore.git && cd manticore/examples/linux
make
# Use the Manticore CLI
manticore basic
cat mcore_*/*0.stdin | ./basic
cat mcore_*/*1.stdin | ./basic
# Use the Manticore API
cd ../script
python count_instructions.py ../linux/helloworld
Option 1: Perform a user install (requires ~/.local/bin
in your PATH
).
echo "PATH=\$PATH:~/.local/bin" >> ~/.profile
source ~/.profile
pip install --user manticore
Option 2: Use a virtual environment (requires virtualenvwrapper or similar).
pip install virtualenvwrapper
echo "source /usr/local/bin/virtualenvwrapper.sh" >> ~/.profile
source ~/.profile
mkvirtualenv manticore
pip install manticore
Option 3: Perform a system install.
sudo pip install manticore
Once installed, the manticore
CLI tool and Python API will be available.
For installing a development version of Manticore, see our wiki.
If you'd like to use redis for state serialization (instead of disk), install
redis using your host package manager, then install manticore as above, but
with [redis]
appended to the name of the package, e.g.
pip install manticore[redis]
Note that this does not make manticore use redis automatically, and you'll still have to manually set the workspace to the redis URI.
$ manticore ./path/to/binary # runs, and creates a mcore_* directory with analysis results
$ manticore ./path/to/binary ab cd # use concrete strings "ab", "cd" as program arguments
$ manticore ./path/to/binary ++ ++ # use two symbolic strings of length two as program arguments
or
# example Manticore script
from manticore import Manticore
hook_pc = 0x400ca0
m = Manticore('./path/to/binary')
@m.hook(hook_pc)
def hook(state):
cpu = state.cpu
print 'eax', cpu.EAX
print cpu.read_int(cpu.ESP)
m.terminate() # tell Manticore to stop
m.run()
Further documentation is available in several places:
-
The wiki contains some basic information about getting started with manticore and contributing
-
The examples directory has some very minimal examples that showcase API features
-
The manticore-examples repository has some more involved examples, for instance solving real CTF problems
-
The API reference has more thorough and in-depth documentation on our API