// the take
A hand-maintained README listing real-world smart contract fuzzing suites (Echidna, Medusa, Foundry invariant tests) from protocols like Aave, Uniswap V4, Euler, and MakerDAO, with links straight to the test code. It's for smart contract security engineers and auditors who want working examples of invariant/fuzz harnesses instead of toy tutorials.
Every entry links directly to the actual test directory in the target repo, not a blog post about it, so you can read real harness code, handler contracts, and invariant definitions from firms like Trail of Bits, Recon, and Enigma Dark. Coverage spans nearly a decade (2018-2025) and includes both Echidna/Medusa and Foundry invariant styles, which is useful for comparing how harness design evolved. Entries include author/firm attribution, which lets you follow specific fuzzing engineers' work across multiple protocols.
It's just a markdown table with no tooling, no categorization by DeFi primitive (AMM vs lending vs bridge) or by bug class found, so finding a relevant example still means opening a dozen links. Many linked branches are feature/audit branches (e.g. 'enigmadark-invariant-suite', 'wf/feat/recon-invariants') that get deleted or rebased after the audit ships, so a chunk of these links will eventually 404 with no archival copy. Several entries just say 'Unknown' for author, and there's no indication of which campaigns actually found bugs versus which just exist as test suites, so it conflates 'has fuzzing' with 'fuzzing was effective'.
We read the README, the directory tree and a key source file in October 2026. Nobody pays to be listed. How we judge →