Claude warns about API key safety, then reads it from .env

User shares anecdote where Claude AI warns against giving API keys to agents, then reads the key from a .env file. The post highlights the irony of AI safety warnings in practice.