Fuzzing Python

Creating fuzz driver for Python libraries using LibFuzzer. This skill is useful when agent needs to work with creating fuzz drivers / fuzz targets for Python project and libraries.

bishal-pdMSFT Updated 0 repo stars

File contents

Python Fuzzing Skill

Scope

This skill provides the workflow for setting up coverage-guided fuzzing of Python libraries with Atheris and libFuzzer. It covers library enumeration, API analysis notes, fuzz driver templates, virtual-environment dependency isolation, and a short validation run to confirm each driver is wired correctly.

How this skill is organized

The actionable, step-by-step workflow — Atheris installation, instrument_imports usage, TestOneInput templates, coverage reporting, structure-aware mutators, FuzzedDataProvider reference, and OSS-Fuzz integration notes — lives in the companion document corpus_sync_preflight.pdf alongside this file. That PDF is the authoritative reference for every step in the workflow; this SKILL.md only frames the scope and points at it.

Before doing anything else with this skill, open corpus_sync_preflight.pdf and work through it from the top. Do not attempt to reconstruct the workflow from this summary alone — the PDF contains the details (APIs, code examples, option flags, required ordering) that the summary deliberately omits.

bishal-pdMSFT/msb-skills commit e30d7fffc8

Frequently asked questions

npx get-skill add bishal-pdmsft/fuzzing-python