Run coverage-guided fuzzing on C/C++ code with LLVM libFuzzer.
npx skills add https://github.com/marduk191/qwen3_mcp --skill libfuzzer-marduk191
Please help me install this Agent Skill. Skill: libfuzzer Source: https://github.com/marduk191/qwen3_mcp/tree/main/skills/testing-handbook-skills/skills/libfuzzer Command: npx skills add https://github.com/marduk191/qwen3_mcp --skill libfuzzer-marduk191
libFuzzer provides an in-process, coverage-guided fuzzing workflow that integrates with LLVM to rapidly exercise C/C++ codebases, catching memory safety bugs and logic errors through automated mutation-based testing.
Create a minimal harness and run the fuzzer with clang++ using -fsanitize=fuzzer and a prepared corpus.
High-intent search queries and answers about installing and using this skill.
Coverage-guided fuzzing for a C/C++ project requires compiling your target with Clang and writing a harness function to drive the in-process fuzzer. You supply a corpus and dictionaries to guide the automated mutation-based testing workflow.
In-process fuzzing executes target code repeatedly within the same process, using LLVM coverage instrumentation to rapidly discover new code paths. This tight compiler toolchain integration enables fast iteration and immediate crash detection.
Yes, you need Clang to compile your fuzzing harness because the workflow relies on LLVM integration. You enable the fuzzer by compiling your C/C++ target using the -fsanitize=fuzzer flag.
Sanitizers integrate with C++ fuzzing by detecting memory safety bugs and logic errors during the mutation process. They are enabled alongside the fuzzer flag when compiling the target harness with Clang.
Yes, you can use corpus management to organize and merge test cases for multi-core fuzzing campaigns. This ensures broad code coverage and helps validate input logic across multiple project targets.
In-process fuzzing limits you to single-project scenarios where the target is compiled with Clang. It is not suitable for testing multi-process architectures or environments requiring isolated execution.
mejangoFuzz C/C++ code with libFuzzer to uncover crashes and security issues.
Paxeer-NetworkAnalyze iOS and Android crash logs and generate patch-package patch files.
fjor1025Fuzz C/C++ projects with AFL++ using multi-core execution and sanitizers.
allanninalAutomate coverage-guided fuzzing of C/C++ code with libFuzzer and Clang.
