Read this alongside section 03, not after it. In this domain security isn’t an advanced topic bolted onto the language — it’s the reason the language is used the way it is.

If you read one note here, read 08-case-studies. It’s the fastest way to understand why everything else in this folder is written the way it is — and its table shows that the money went somewhere different from where the field’s attention goes.

Reading order

  1. why-this-is-different — [Intermediate] — no patching, no perimeter, funded adversaries, composability as attack surface, and what actually reduces risk (ranked)
  2. reentrancy — [Intermediate] — the canonical bug, the three variants people miss (cross-function, cross-contract, read-only), and why .transfer() is not the fix
  3. access-control-and-key-management — [Intermediate] — the largest category of losses: tx.origin, the initialiser trap, multisig signer independence, and the four signature checks
  4. arithmetic-and-rounding — [Intermediate] — 0.8 fixed overflow and not precision; multiply before dividing; round in the protocol’s favour; the vault inflation attack
  5. oracle-and-price-manipulation — [Advanced] — flash loans make capital rentable, so spot price as truth is the field’s most reliable exploit. TWAPs, staleness checks, and not needing a price at all
  6. mev-front-running-and-ordering — [Advanced] — sandwiching, PBS, why MEV can’t be eliminated, and what actually protects users
  7. the-audit-process — [Intermediate] — what an audit is and isn’t, how to prepare, and how to read someone else’s report
  8. case-studies — [Intermediate] — The DAO, Parity, Ronin, Wormhole, Nomad, Euler, Curve, Mango — and the pattern across all of them
↑ Back to Web3 & Blockchain