Skip to main content

Official Certified Edible Dinosaurs CTF toolkit.

Project description


The official Certified Edible Dinosaurs CTF toolkit. pwnypack attempts to provide a toolset which can be used to more easily develop CTF solutions.

Build Status


After seeing the excellent pwntools by Gallopsled, I got interested in building my own CTF toolkit. pwntools is much more complete so you should probably use that. pwnypack was created mostly out of curiosity.


First, make sure your setuptools and pip packages are up to date:

$ pip install -U setuptools pip

To install the latest released version of pwnypack with all optional dependencies, run:

$ pip install --no-binary capstone,keystone-engine pwnypack[all]

To install the latest released version of pwnypack with minimal dependencies, run:

$ pip install pwnypack

Other available install targets are:

  • --no-binary keystone-engine pwnypack[asm] - installs keystone-engine to support all assembler targets without any additional depenencies.

  • --no-binary capstone pwnypack[disasm] - installs capstone for AT&T and intel syntax disassembly, required to disassemble ARM binaries.

  • --no-binary capstone pwnypack[rop] - installs capstone to validate ROP gadgets.

  • pwnypack[ssh] - installs paramiko to enable the Flow module to connect to SSH servers.

  • pwnypack[shell] - installs ipython to support the enhanced pwnypack REPL environment.

  • pwnypack[pwnbook] - installs jupyter to support the pwnbook jupyter notebook.

If you want to use the interactive shell I highly recommend installing either bpython or ipython as those packages can make your time in the shell a lot more enjoyable.


You can also use our published docker images.

To start an ipython powered pwnypack shell:

docker pull edibledinos/pwnypack:latest
docker run --rm -it edibledinos/pwnypack:latest

Or, to run pwnbook:

docker pull edibledinos/pwnbook:latest
docker run --rm -it -p 8888:8888 edibledinos/pwnbook:latest

Both images expose a volume (/projects). Feel free to mount something interesting there.

Three tags are available:

  • py3 (or: latest) installs python 3 and pwnypack/pwnbook.

  • py2 installs python 2 and pwnypack/pwnbook.


To import all of pwnypack into your global namespace, use:

>>> from pwny import *

Or, if you’re using python 2.7+ or python 3.3+, try the customized bpython or IPython shell:

$ pwny shell

If you have bpython and/or IPython installed you can use --bpython, --ipython or --python to select which interactive kernel to use.

I promise that effort will be put into not exposing unnecessary stuff and thus overly polluting your global namespace.

For an example, check out the Big Prison Fence example in the wiki.

Common errors

Capstone/keystone fails to import the dynamic library.

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/ingmar/.virtualenvs/pp/lib/python3.5/site-packages/capstone/", line 230, in <module>
    raise ImportError("ERROR: fail to load the dynamic library.")
ImportError: ERROR: fail to load the dynamic library.

The capstone and keystone packages have a bug which when used with a new verion of pip will end up installing the dynamic libraries in the wrong location on some platforms. Re-install capstone and/or keystone using:

$ pip install --no-binary capstone capstone
$ pip install --no-binary keystone-engine keystone-engine

SyntaxError when importing pwnypack.

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "pwny/", line 9, in <module>
    from pwnypack.pwnbook import *
  File "pwnypack/", line 2, in <module>
    from jupyter_client import kernelspec as kernelspec
  File "/Users/ingmar/.virtualenvs/pwny26/lib/python2.6/site-packages/jupyter_client/", line 4, in <module>
    from .connect import *
  File "/Users/ingmar/.virtualenvs/pwny26/lib/python2.6/site-packages/jupyter_client/", line 23, in <module>
    from traitlets.config import LoggingConfigurable
  File "/Users/ingmar/.virtualenvs/pwny26/lib/python2.6/site-packages/traitlets/", line 1, in <module>
    from .traitlets import *
  File "/Users/ingmar/.virtualenvs/pwny26/lib/python2.6/site-packages/traitlets/", line 1331
    return {n: t for (n, t) in cls.class_traits(**metadata).items()
SyntaxError: invalid syntax

You’ve installed jupyter notebooks on python 2.6. Use a more modern version of python.


pwnypack’s API documentation is hosted on readthedocs.

For information on the commandline apps use the built in help function:

$ pwny --help
$ pwny shell --help


pwnypack was created by Certified Edible Dinosaurs (dsc & doskop). If you want to contribute, feel free to fork and create a pull request on GitHub.

Current contributors:

  • blasty <> contributed the ARM shellcode generator.


pwnypack is distributed under the MIT license.

Release history

next release

0.9.0 (2017-11-17)

  • Adding missing dev packages on docker images.

  • Docker images clone latest branch of pwnypack repository.

  • Fix magic command registration on IPython >= 5.0.0.

  • Support parsing (a subset of) ~/.ssh/config in flow.SSHClient.

  • Fix aarch64 support when using keystone engine.

  • Added padding oracle attack functions.

  • Added keystone-engine dependency for ‘asm’ and ‘all’ targets.

  • Make pwnbook, shell and ssh targets require python 2.7+.

  • Added internals for python 3.5.2 and 3.6.

0.8.0 (2016-05-17)

  • Return empty list when trying to read non-existing .dynamic section.

  • Don’t print newline when piping the output of a gadget.

  • Fix output of raw binary data on python 3.

  • Add pwnypack extension for ipython.

  • Add pwnypack jupyter notebook wrapper (pwnbook).

  • Moved and renamed util.pickle_call to pickle.pickle_invoke.

  • Added pickle_func that pickles a function and its invocation.

  • Added support for using GNU binutils to assemble AT&T and intel syntax.

  • Added support for assembling/disassembling ARM using binutils/capstone.

  • Use extras_require to make capstone, paramiko and jupyter optional.

  • Add Dockerfile for pwnypack shell and pwnbook.

  • Fix interact on python 3 in Flow.

  • Add python bytecode manipulation functions.

  • Added shellcode generator for X86/X86_64, ARM (+Thumb) and AArch64.

  • Use keystone engine as assembler engine by default.

  • Added xor mask finder.

  • Added python independent marshal and .pyc loader.

  • Fix internal escaping of reghex expressions.

  • Allow wildcards when searching for ROP gadgets using assembly statements.

0.7.2 (2016-03-11)

  • Added support for .dynamic section parsing to ELF class.

  • Added checksec command line app.

  • Make pwnypack available as a universal wheel.

0.7.1 (2016-03-07)

  • Add support for bpython and plain python interactive shells.

  • Fix missing newline after the output of several commandline apps.

  • Added PHP serialized data generation function and helper.

  • Add enurlform / deurlform / enurlquote / deurlquote functions.

0.7.0 (2015-07-12)

  • Made IPython an optional dependency (pip install pwnypack[shell]).

  • Added pickle_call function to easily execute a function on unpickle.

  • Added format string vulnerability exploit builder.

  • Renamed TCPSocketChannel to TCPClientSocketChannel.

  • Added TCPServerSocketChannel which can listen for an incoming connection.

  • Added Flow.interact() method.

  • Added support for connecting to SSH servers from Flow.

0.6.0 (2015-04-14)

  • Bugfixes (and travis-ci integration).

  • API documentation and docstrings.

  • Cycle-find can read from stdin.

  • Major refactoring of ELF class. It can now parse section headers, program headers, symbol tables and extract section, symbols.

  • Major refactoring of Target class. It’s no longer tied to ELF (ELF is still a subclass of Target though).

  • A reghex compiler.

  • Verifying ROP gadget finder.

  • Disassembler functionality (based on ndisasm or capstone).

  • The ability to redirect stderr to stdout in flow.ProcessChannel.

  • The ability to create symlinks for commandline apps.

  • New commandline apps:
    • asm to assemble from commandline.

    • symbols to list the symbol table of an ELF file.

    • gadget to find ROP gadgets in an ELF file.

    • symbol-extract to extract a symbol from an ELF file.

    • symbol-disasm to disassemble a symbol in an ELF file.

0.5.2 (2015-03-22)

  • Added command line apps and a customized IPython shell.

0.5.1 (2015-03-21)

  • Python3 fixes for flow:
    • Use latin1 for echo mode as not everything will be encodable as utf-8.

    • Disable buffering on subprocess.

0.5.0 (2015-03-21)

  • Initial release.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pwnypack-0.9.0.tar.gz (89.2 kB view hashes)

Uploaded Source

Built Distribution

pwnypack-0.9.0-py2.py3-none-any.whl (106.8 kB view hashes)

Uploaded Python 2 Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page