Privacy: are these prompts sent to Anthropic when I run them?

Yes, that’s how Claude works. When you paste one of our prompts into Claude Code, it’s sent to Anthropic’s API along with whatever code Claude reads from your repo. However: Anthropic’s commercial API (Claude Pro / Team / API plan) does not train on your inputs, that’s contractual under their Commercial Terms. You can also enable zero-data-retention on Enterprise plans. The prompts themselves in our library are public-domain MIT, so there’s nothing confidential to leak by sharing the recipe.