ossfuzz

To enroll an open-source project for automated fuzzing, you need to create the required OSS-Fuzz configuration files: project.yaml, Dockerfile, build.sh, and helper.py. These files define the build environment and fuzzing harnesses for continuous testing.

Fuzzing harnesses are fully supported with Rust and Go projects, alongside C/C++ and Python. The system uses sanitizers like ASan and UBSan to build and run these harnesses, enabling crash discovery across multiple programming languages.

Crash triage involves using helper.py to reproduce the crashes discovered by the fuzzers. After running tests with sanitizers, the system collects crash reports and coverage data to help you review and resolve security issues.

Coverage reports for a C++ fuzzing target are generated using the integrated coverage tooling. By running the fuzzing harnesses and collecting the resulting coverage data, the system produces detailed reports for code improvement analysis.

Docker is required to configure continuous fuzzing infrastructure because the enrollment process uses a Dockerfile to define the build environment. This containerized setup ensures reproducible builds and consistent fuzzer execution.

Distributed fuzzing for open-source software requires maintaining specific configuration files like project.yaml and build.sh. Limitations include the need for manual harness development and ensuring your project builds successfully within the Docker environment.