Fuzzers like AFL and GoFuzz are amazing pieces of security engineering, but were clearly built to be run in an ad-hoc way. Some other attempts to help alleviate this problem exist (like afl-fuzz-docker and Cloudfuzzer), but here at Coinbase we needed something that would scale within AWS to suit our needs and not create lots of extra busywork for developers.
Taking some cues from Google's ClusterFuzz, we built Maxfuzz - a fuzzing framework that abstracts the annoying and tedious bits of running a fuzzing campaign away and makes it easy to deploy fuzzers, collect results, and get back to the other important things you're doing.
This guide assumes you have Docker, docker-compose and make installed. For more on getting set up, please check out our Development Environment Setup wiki page. Most of the bootstrapping and deployment process is done through a Makefile for simplicity and repeatability.
First, build the base Maxfuzz Docker image: make build
Then spin up a local instance of Maxfuzz, running some sample fuzzers: make deploy-dev
This launches a basic fuzzer that fuzzes some vulnerable C code, which can be found in ./fuzzers/vulnerable/.
Upon deploying the fuzzer locally you should see some crashes appear fairly quickly in ./sync/afl-vulnerable/crashes/.
As these crashes are uncovered by AFL, you should see corresponding logs from the Docker containers in the docker-compose view.
For more details on how to use Maxfuzz, please check out our How to Use Maxfuzz wiki page
Contribution to this project is extremely welcome!
Please direct all Issues/Pull Requests to the Bleeding Edge Repository - the Coinbase Repository will pull stable releases from there.
To get started, check out our wiki pages on: