Who can still change this contract?

Paste-free answers for 40 widely-used contracts. For each one: can the code be replaced, who holds the key that replaces it, what the privileged role can still do to holders — and whether a live bug bounty pays for finding a hole in it. Every claim links to the evidence it came from.

Why this exists: a block explorer shows you a storage slot. It does not tell you that the address in it is a plain wallet, that the wallet can mint, and that a live program pays six figures for a bug in the same code. Those facts sit in four different places. Here they sit in one.

Reading the Powers column: the first number is capabilities the contract’s own published interface declares. The second, marked +n?, is selectors found in the compiled code that were not confirmed by a source — a swap router shows mint there because it calls mint on a pool, not because it can mint. The two are never added together.

The whole table as data

Free, no key, CC0: /c/index.json — and one file per contract at /c/<chain>/<address>.json with the full evidence. For an address that is not on this list, a live lookup is $0.05.

What this is not

It is not an audit and there is no safety score here — deliberately. A score invites you to stop reading, and the interesting part is always which power, held by whom. Upgradeable is not a synonym for dangerous: most large protocols are upgradeable on purpose. What matters is who holds the key and what stands between them and your balance.