Nick Perry shows five prompt modifiers that change how Claude responds: Truth mode stops evasive answers, /human generates natural prose, Red team plays devil's advocate, ELI-10 simplifies for a 10-year-old, and Future answers as your wiser self would.
Here are 5 Claude codes that completely change how it talks to you. 1. Truth mode. Put it before your prompt and Claude stops gaslighting you. 2. /human. Put it before your prompt and you get text that actually sounds like a person wrote it, not a robot. 3. Red team. Type this and Claude plays devil's advocate against your own ideas. It argues the other side until you're actually smarter. 4. ELI-10. Drop it anywhere and Claude explains the most complicated thing in terms a 10-year-old would get. 5. Future. Add this and Claude answers the way your future self would, the version of you that already figured it out.
