25 Claude Hacks for Power Users
From Wikiprompt, the free prompt encyclopedia
25 Claude Hacks for Power Users A comprehensive list of 25 practical tips for using Claude effectively, covering app usage, skills, projects, prompt techniques, and workflow optimization.
Prompt ContentSave
π
1. Use the Claude Desktop app. The browser is the demo. The app is where Cowork lives.
2. Use Cowork, not the chatbot. If you're typing into a chat box, you're using it like it's 2025.
3. Forget .md files & folders. Folders leak old outputs into your context, and nobody on Earth maintains those files weekly.
4. Skills and Projects. That's the entire system now. A skill is a capability you teach once. A project is a place you go back to.
5. Delete your about-me file. Rebuild it as a skill so it follows you into every single chat.
6. Turn every setting off. Empty your global instructions. Leave Connectors on. That's it.
7. Projects remember between chats. And you can share one with your entire team in two clicks.
8. The longer your chat, the dumber it gets. Anthropic's own engineers said it out loud.
9. So stop typing "no, I meant...". Click "Restart the conversation from here" on an earlier message.
10. A token is roughly one word. Every word you send & receive is money. Message 30 costs you 31 times message 1.
11. Point Fable 5 at your single hardest problem. Two turns. Then switch models. Turns are the most expensive habit you have.
12. Opus 5 High for everything else. High, not Max.
13. Don't give Claude steps. Give it the goal & get out of the way. Steps drag it down to old Claude.
14. Don't paste logins, passwords, or keys. Ever.
15. Skills fire themselves. You don't prompt them. You type /command.
16. It will sound 100% sure when it's dead wrong.
17. Don't trust agreement either. Claude is a sycophant. Pleasing you and being right are two different jobs.
18. Use Research mode (the "+" in the chat box). Almost nobody knows it's in there. Give it something genuinely hard.
19. Add "use AskUserQuestion" to your prompt. Claude interviews you before it answers. It prompts itself better than you ever will.
20. Talk, don't type. Cowork reads 100,000 words in 15 secs, then waits for you at 60 words per minute. You speak at 150.
21. Ask for five very different versions. Then pick one. That part is taste, and taste is the part it can't do for you.
22. Make it audit its own answer before you trust a single line.
23. Feed it too much context and it goes in circles. Same answers every time, zero creativity. Keep the context in the project and the chat clean.
24. The first draft is yours to fix, not to ship. Always.
25. The only rule that matters: outsource the thinking, never the understanding.
Sign in to see the full prompt
Continue with:
By logging in, you agree to our Terms of Use and Privacy Policy
Usage
This prompt is designed for use with productivity. Copy the prompt content above and paste it into your preferred AI tool.
For best results, you may customize the placeholders (indicated by square brackets or capital letters) with your specific requirements.
References
- Category: productivity Prompts
- Source: https://x.com/rubenhassid/status/2081696106988073143
Talk
0 comments