<aside> 📌 You're using Claude Code at maybe 20% of its power. These 7 commands unlock the other 80%.

</aside>

Code 1: ultrathink

Type "ultrathink" anywhere in your prompt and Claude shifts into deep reasoning mode. It thinks 10x harder before answering — analyzing edge cases, considering alternatives, and actually working through the problem instead of rushing.

• Just type the word "ultrathink" in your message
• Use it for complex architecture decisions, tricky debugging, anything important
• Don't use it for simple questions — it's overkill for "rename this variable"

Code 2: /voice

Stop typing. Hold the space bar and just talk to Claude. It transcribes in real time and responds like normal. Talking through a problem is 5x faster than typing it out.

• Hold spacebar to talk, release to send
• Perfect for brainstorming, explaining complex problems, or when your hands are tired
• Type /voice to toggle it on

Code 3: /compact

Long conversations make Claude forget things. /compact compresses your entire chat without losing the important context. Claude stays sharp, your session stays fast.

• Run it when Claude starts repeating itself or feeling sluggish
• Keeps all the important context, drops the noise
• Same session, same project — just tighter

Code 4: /schedule

Set tasks on a timer. "Run this every morning at 6AM." It executes in the cloud — even when your computer is off. Set it and forget it.

• Runs on Anthropic's cloud infrastructure
• Survives reboots, sleep mode, everything