Etherscan has been in Claude's directory since September. 16 tools, 60+ EVM chains, no API key, read-only. Feature list is fine. Deployer, verified source, ABI, logs, gas, all from one prompt. If you check fresh contracts all day, that's a lot of explorer tabs gone. Now the docs. There's a heading "Guard against prompt injection" and one sentence under it: be careful when you combine Etherscan with other connectors in the same chat. The tool only reads. It never signs, never moves funds. So what is it guarding against? My read, not their claim: text onchain is writable by anyone. Token name, symbol, comments inside verified source. A scam deployer controls that text, and your model reads it as part of the answer. Now put Drive or Gmail in the same chat and think about who else is reading it. frens, I'd keep Etherscan in a clean chat with nothing attached and check fresh deploys there. Do you run your explorer tools separate from mail and docs, or does everything live in one chat? Next post I run it on a junk contract and show what comes back.