wtf is frostyverify-cookbook?

FrostyVerify Cookbook is a collection of copy paste recipes for connecting a service called FrostyVerify, which checks whether email addresses are real and safe to send to, into different tools people already use: Claude Code, an outbound sales pipeline, Slack, and Google Sheets. FrostyVerify itself is a separate hosted product, and this repository is not its source code, it is a set of guides for wiring it into other systems. The core idea behind the project is that sending emails to addresses that bounce, are spam traps, or are disabled teaches email providers that you do not know who you are emailing, which over time hurts how many of your emails reach the inbox at all. Verifying a list before sending is meant to be cheap insurance against that. Because FrostyVerify offers both an API and something called an MCP server, an AI coding assistant can run the verification step automatically instead of a person doing it as a manual chore. Setup is a single command that registers FrostyVerify as an MCP tool inside Claude Code, using an API key you generate on the FrostyVerify website. Once connected, Claude Code gains four tools: one to verify a single email, one to verify a list in bulk, one to specifically check catch-all addresses, and one to check the status of a bulk verification job. New accounts get 500 free verification credits, enough to try every recipe in the repository. The repository lists six specific recipes: verifying a CSV file of leads directly from Claude Code, building a self-cleaning outbound pipeline that scrapes, deduplicates, and verifies leads automatically, running verification from within Slack by dropping in a CSV, verifying a column of emails inside Google Sheets using a custom formula, a single prompt that teaches any AI assistant how to use the API without installing anything, and a recipe for recovering some of the catch-all addresses that other tools simply discard. Email results are grouped into categories such as safe, role, catch-all, disposable, spam trap, and invalid, and the README states that only the safe and role categories should be treated as safe to send to. The project also documents the raw API directly, showing how to verify one email, verify a list asynchronously, and poll for job completion using plain HTTP requests. It is built by a company called Frostmailer, which works on cold email infrastructure.