fuzztest

tsz for me is an experiment to see how can this kind of work be done better. With a slight difference that tsz is not a direct port and it's a different architecture. I'm also not claiming to have answers but I've learned a ton. A few things that works - Test before code, Bun had lots of test so that's good but maybe they could start by asking Mythos to write like 20k additional tests that pass on Zig Bun first. - Deterministic anti-slop features. LLMs love to solve the problem in the wrong abstraction layer or place. There are many ways to catch this with deterministic tests. I do this in tsz a lot - Roadmap that constantly evolving by humans. - Taking a pause and looking how the progress is going and undoing slop - Fuzztest(https://github.com/google/fuzztest) style "trying to break things" with the powers of LLM

google/syzkaller @main: https://github.com/google/syzkaller But there is also fuzztest? google/fuzztest: https://github.com/google/fuzztest

> get the impression that AFL is considered better. The authors of llvm fuzz stopped working on it in favour of some other thing, which they then stopped working on in favour of https://github.com/google/fuzztest Thanks, this kind of social stuff can be useful -- it looks like all the resources folks shared seem to favor AFL.

LLVM ships with a fuzzing library, docs at https://llvm.org/docs/LibFuzzer.html. I get the impression that AFL is considered better. The authors of llvm fuzz stopped working on it in favour of some other thing, which they then stopped working on in favour of https://github.com/google/fuzztest, which seems to be broadly useless as a fuzzer implementation. But whatever, the llvm fuzzer lives on and has uses in tree and occasional updates. I found it much easier to get started with than AFL. I wrote a program that takes a byte array as input and drives the library under test with it, attached that to llvm's fuzzer and left it running. You end up with a lot of files containing some bytes that did something vaguely interesting with the program. Good experience overall. You might get some meaning out of https://github.com/JonChesterfield/bigint