@penzlik
Etherscan has been in Claude's directory since September. 16 tools, 60+ EVM chains, no API key, read-only.
One thing from their docs: a heading "Guard against prompt injection" and a single line to be careful stacking it with other connectors in one chat.
My read, not their claim: token names and source comments are text anyone can write onchain, and your model reads it.
Etherscan in its own clean chat, or all connectors in one?
0 reply
0 recast
2 reactions