Claude Code Daily: 5 Commands That Do the Real Work
Get the exact Claude Code slash commands and flags used daily, with the real-world patterns that save hours per week.
Free checklist · copy-paste ready · built by the BrainVaultAI team
Claude Code is a CLI, not a chat interface. The people getting 10x results from it have a small set of commands they run every day without thinking. This is my actual list. Not a tour of every feature. Just the five I reach for constantly and couldn't work without. Each entry includes the command, what it actually does, and when to use it vs. when not to.
Command 1: /init (Start Smart, Not Blank)
Command: /init What it does: Reads your codebase and writes a CLAUDE.md file that gives Claude persistent context about your project, stack, conventions, key files, constraints. When to use: First thing in any new project. Also re-run it after a major refactor when the codebase has drifted from the original CLAUDE.md. When NOT to use: Don't rely on it as a substitute for writing a real CLAUDE.md. It gives you a starting draft, edit it aggressively before trusting it.
Command 2: claude --continue (Preserve Context Across Sessions)
Command: claude --continue What it does: Resumes your most recent conversation instead of starting fresh. Keeps the full context of what you were working on. When to use: Every time you come back to a task you started in a previous session. Eliminates the 'here's what we were doing' re-briefing. ProTip: Combine with /resume-session if you have that skill installed. The two together mean you never lose thread on a multi-day task.
Command 3: claude -p (One-Shot Pipe Mode)
Command: claude -p "<prompt>" What it does: Runs Claude non-interactively and returns output to stdout. Pipeable. Scriptable. Works in cron jobs, shell scripts, and CI. When to use: Whenever you want Claude as a step in a pipeline. Examples: pipe a log file in and ask for anomalies, pipe a diff in and get a plain-English changelog, run a nightly analysis job. Example: git diff HEAD~1 | claude -p "Summarize these changes in plain English for a non-technical stakeholder."
Command 4: /clear (Surgical Context Reset)
Command: /clear What it does: Wipes the conversation context without ending the session. Claude starts fresh on the next message. When to use: When you've finished one task and are switching to something unrelated. Also use it when Claude starts getting confused or losing coherence, context buildup is often the culprit. When NOT to use: Don't use it mid-task. If you're in the middle of something complex, /clear will cost you more time re-briefing than you'd save.
Command 5: /cost (Track Before It Bites You)
Command: /cost What it does: Shows the token usage and estimated cost for the current session. When to use: Before long agentic runs. After any session that felt expensive. Whenever you're debugging why a workflow costs more than expected. ProTip: Run /cost at the start of your day and at the end. The delta tells you where your token budget is going. Most people are shocked the first time they see it.
Want the editable version?
We'll send the editable version of this asset, plus a short series of our highest-leverage AI plays for operators. No spam. Unsubscribe in one click.
Free. No spam. Unsubscribe in one click. By submitting you agree to receive emails from BrainVaultAI.