Fuzz pure Python code and Python C extensions with coverage-guided libFuzzer instrumentation.
npx skills add https://github.com/TECH-HY/SKILLS --skill atheris-tech-hy
Please help me install this Agent Skill. Skill: atheris Source: https://github.com/TECH-HY/SKILLS/tree/main/skills/atheris Command: npx skills add https://github.com/TECH-HY/SKILLS --skill atheris-tech-hy
atheris.instrument_func and atheris.instrument_imports() to generate coverage-guided random inputs. - C Extension Fuzzing: Compile native extensions with clang and AddressSanitizer flags, configure LD_PRELOAD, and detect memory corruption in libraries like cbor2. - Corpus and Campaign Management: Create seed corpora, minimize them with libFuzzer merge mode, and run parallel fuzzing campaigns with workers and time limits. - Use Case: You maintain a Python library that parses untrusted input. Write a harness that feeds random bytes to your parser, run it in the provided Docker environment with ASan enabled, and catch buffer overflows and unexpected exceptions before release. ## Quick Start Write a fuzzing harness for my Python parser function using Atheris and show me how to run it with a seed corpus.High-intent search queries and answers about installing and using this skill.
Write a harness defining a test_one_input(data) function that calls your target code, decorate it with @atheris.instrument_func, then call atheris.Setup(sys.argv, test_one_input) and atheris.Fuzz(). Run the script with python fuzz.py, optionally passing a corpus directory.
Atheris is a coverage-guided fuzzer that mutates byte inputs to explore code paths and find crashes, including memory corruption in C extensions. Hypothesis is a property-based testing library that generates type-aware inputs to verify code invariants. Choose Atheris for crash discovery and Hypothesis for verifying logical properties.
Yes, Atheris supports fuzzing C extensions compiled with clang using -fsanitize=address,fuzzer-no-link flags. You must set LD_PRELOAD to the asan_with_fuzzer.so library and install the extension from source with --no-binary so it compiles with instrumentation.
This happens when LD_PRELOAD is not set to the asan_with_fuzzer.so shared library. Export LD_PRELOAD pointing to that file inside the atheris package directory, or use the provided Docker environment where it is preconfigured.
Atheris supports 32-bit and 64-bit Linux and macOS, with Linux recommended for simpler dependency management and faster execution. It requires Python 3.7 or later and a recent clang version, and a Docker setup is provided for a preconfigured environment.
tottenjordanWrite fuzzing harnesses for libFuzzer, AFL++, cargo-fuzz, and go-fuzz targets.
dslsdzcReproduce, minimize, and root-cause program crashes using ASAN, gdb, rr, and afl-tmin.
monmacllcappAutomates C/C++ fuzzing with libFuzzer, including setup, compilation, and execution.
tottenjordanGuides multi-core fuzzing of C/C++ projects using AFL++ instrumentation and campaign management.