#Claude

Let Claude write your prompt: the meta-prompting loop

The fastest way to a better prompt is to have Claude help you write it. Here is the draft, critique, refine, run loop I use to get sharper results with less thrash.

Use Claude Code remotely from your phone or the web

Step away from your desk without stopping a Claude Code run. Turn on remote control and pick the session up from your phone, claude.ai/code, or another machine.

Stop running one giant chat: manage context in Claude Code

A long-running Claude Code session gets slower, pricier, and less accurate. How I keep context lean, and what a big context window actually costs you.

/clear, /compact, or a new session: when to reset Claude Code

Claude Code gives you two ways to trim a session and a third option to start over. /clear wipes, /compact summarizes, and a simple rule tells you which one the moment calls for.

Plan before you build: getting the most out of plan mode

Plan mode makes Claude Code lay out its approach and wait for your sign-off before it touches a file. I use it to skip the expensive redo loop.

Sub-agents in Claude Code: smart model, cheap workers

Let a capable model plan and cheap models do the grunt work. How Claude Code sub-agents work, how to set one up, and the token math that makes them pay off.

Write a CLAUDE.md that stops you repeating yourself

A good CLAUDE.md is project memory: say the durable facts once and Claude Code loads them every session. What to put in it, what to leave out, and why it saves both tokens and mistakes.

Which Claude model should you use in Claude Code

Opus, Sonnet, Haiku, or Fable. A plain comparison of the Claude models by cost, speed, and capability, plus which one I reach for in Claude Code and when.

Context rot: why a long Claude Code chat gives worse answers

Claude Code does get worse as a session grows, and it is not your imagination. Here is what context rot is, why it happens, and why pasting everything backfires.