Every Claude Command You Need to Know in 2026 (89 Commands Complete Guide)

Every Claude Command You Need to Know in 2026
Complete Reference Guide · 89 Commands

Every Claude Command
You Need to Master

Most people type a question into Claude and call it a day. But beneath the surface lies a structured system of 89 commands — organized across 11 categories — that can transform how you think, write, build, and research. This guide covers all of them, plus a practical adoption roadmap.

📅 May 2026 ⏱ 10 min read 🧠 89 Commands Documented ✦ Beginner to Advanced
89
Total Commands
11
Categories
10x
Productivity Boost
5
Bonus Shortcuts

Why Commands Change Everything

Claude is not just a chatbot — it is a structured reasoning engine that responds best when given clear, directive prompts. The 89 commands in this guide are not magic words; they are mental frameworks that help you communicate intent precisely. When you prefix a request with /analyze, you are signaling: “I want this broken down systematically.” When you use /challenge, you are asking Claude to push back. The result? Better outputs, fewer revisions, and conversations that actually go somewhere.

The shift in mindset: Stop thinking of Claude as a search engine where you ask questions. Start thinking of it as a highly capable collaborator who responds to clear direction. Commands are that direction.

The Complete Command Reference

01 Start & Create
/newStart a new chat session
/projectCreate a project
/uploadUpload files for reference
/pastePaste content from clipboard
/templateUse a reusable template
/importImport from a file
/scanScan documents for key info
/voiceUse voice input mode
02 Focus & Context
/focusSet the main objective
/contextAdd background information
/detailsProvide more detail
/examplesGive illustrative examples
/clarifyAsk follow-up questions
/defineDefine specific terms
/assumptionsList working assumptions
/prioritiesSet priorities for the task
/constraintsDefine constraints upfront
03 Think & Solve
/analyzeBreak a problem down
/compareCompare multiple options
/pros-consList pros and cons
/evaluateEvaluate ideas or plans
/recommendGet clear recommendations
/brainstormGenerate many ideas fast
/solveWork through the problem
/challengeChallenge assumptions
04 Write & Edit
/writeGenerate new content
/editEdit for clarity
/rewriteRewrite more effectively
/shortenMake it more concise
/expandAdd more detail
/improveImprove the writing overall
/summarizeSummarize the content
/paraphraseParaphrase in new words
/proofreadProofread for errors
05 Organize & Structure
/outlineCreate a structured outline
/structureOrganize messy content
/bulletConvert to bullet points
/numberedMake a numbered list
/tableCreate a comparison table
/summarySummarize content cleanly
/key-pointsExtract the key points
/mindmapCreate a mind map layout
/flowchartCreate a flowchart diagram
06 Code & Tech
/codeWrite code from scratch
/debugFix issues in code
/explainExplain code logic
/optimizeImprove code performance
/refactorRefactor for cleanliness
/testWrite test cases
/convertConvert between formats
/documentationWrite technical docs
/reviewReview code quality
07 Data & Analysis
/analyze-dataAnalyze raw data
/visualizeCreate charts & visuals
/insightsExtract useful insights
/forecastMake predictions
/reportGenerate a full report
/statsCalculate statistics
/cleanClean and normalize data
08 Automate & Integrate
/workflowCreate a workflow
/automateAutomate repetitive tasks
/apiWork with APIs
/integrateConnect tools together
/scheduleSet up reminders
/triggerDefine automation triggers
/tasklistCreate a task list
/checklistBuild a checklist
09 Personalize & Control
/preferencesSet your preferences
/memoryManage memory & context
/toneAdjust the output tone
/styleChange writing style
/lengthControl response length
/formatChange output format
/resetReset the conversation
/clearClear context entirely
10 Learn & Research
/searchSearch the web
/researchDeep research on a topic
/learnExplain a topic thoroughly
/tldrGive a TL;DR summary
/sourcesFind credible sources
/fact-checkVerify factual claims
/exploreExplore a topic freely
11 Collaborate & Share
/shareShare the conversation
/exportExport the content
/downloadDownload as a file
/copyCopy to clipboard
/emailDraft email content
/publishPublish content
/feedbackSend feedback

Bonus: Power Shortcuts

These five principles multiply the effectiveness of every command above. Think of them as operating principles, not one-off tips.

  • Use / + command for quick access. On desktop, the slash shortcut opens the command palette instantly — no need to type out natural language instructions from scratch.
  • Combine commands for better results. Stack commands within a single prompt: /analyze this contract, then /recommend three action items, then /email me a summary” — one prompt, three structured outputs.
  • Add context early for better answers. Begin every complex session with /context before any task command. Context-first prompts consistently return more relevant, accurate, and actionable responses.
  • Be specific and clear. Commands work best with specificity. Instead of /write a blog post,” try /write a 600-word blog post for senior developers on API rate limiting, technical tone, three practical examples.”
  • Iterate and refine. Use /improve, /shorten, or /tone after an initial output rather than rewriting the entire prompt. Refinement commands preserve the good parts while fixing what needs changing.

How to Actually Adopt These Commands

Knowing 89 commands is different from using them well. Here is a practical five-stage adoption roadmap that takes you from curious beginner to confident power user.

1

Start with five core commands

Do not try to learn all 89 at once. Begin with the five most universally useful: /write, /summarize, /analyze, /brainstorm, and /edit. Use each one deliberately for a week. Notice how your outputs change compared to freeform prompting.

2

Add context to every session

Make /context a habit before any substantive task. Tell Claude who you are, what the project is, and what a good output looks like. This single change will improve your results more than any other technique. Think of it as briefing a colleague before delegating a task.

3

Stack commands deliberately

Once individual commands feel natural, start combining them. A practical example: open with /context, run /research on a topic, follow with /key-points to extract the core insights, and finish with /write to turn those insights into a draft. One session, four commands, publishable output.

4

Build your personal workflow templates

Identify the three to five tasks you do repeatedly — writing emails, summarizing documents, reviewing code, preparing reports. Build a standard command sequence for each one and save it using /template. Repeatable workflows eliminate decision fatigue and ensure consistent quality.

5

Explore the advanced categories

Once your core workflow is solid, explore the categories that unlock new capabilities: Automate & Integrate (for connecting tools), Data & Analysis (for turning numbers into insights), and Personalize & Control (for tuning Claude to your exact preferences). These are where advanced users find the most leverage.

Quick Reference: The 15 Most-Used Commands

Based on real-world usage patterns, these are the commands that deliver the most value for the broadest range of tasks.

Command Best Used When
/analyzeYou need a problem broken down before taking action
/contextStarting any complex or multi-step session
/summarizeDigesting long documents, reports, or transcripts
/writeGenerating any kind of content from scratch
/brainstormGenerating a wide range of options quickly
/editImproving existing content for clarity and impact
/codeWriting or generating any technical code
/debugDiagnosing errors or unexpected behavior in code
/explainUnderstanding complex concepts or code you didn’t write
/researchDeep-diving into an unfamiliar topic
/pros-consMaking a decision between two or more options
/outlineStructuring a large piece of work before writing it
/toneAdjusting output to match an audience or brand voice
/fact-checkVerifying claims before publishing or presenting
/challengeTesting the strength of an idea or argument

Professional Tips for Getting the Most Out of Claude

🎯
One task per prompt

Keep each prompt focused on a single command or objective. Clarity in input produces clarity in output. Multi-intent prompts produce mixed results.

📐
Specify the format

Use /format or describe your desired output structure explicitly. “Respond in JSON,” “use numbered steps,” or “write in three paragraphs” — format direction matters.

🔁
Iterate, don’t rewrite

When an output is 70% there, use /improve or /shorten rather than starting over. Refining is faster than re-prompting from scratch.

🗂️
Use Projects for ongoing work

The /project command keeps context persistent across sessions. For anything spanning multiple days, a project retains history and avoids repetitive re-briefing.

⚖️
Challenge before committing

Before acting on any plan or decision, run /challenge. Ask Claude to argue against your own plan. The strongest ideas survive this; weaker ones improve from it.

🔍
Verify critical outputs

Use /fact-check and /sources on anything time-sensitive or high-stakes before publishing. AI outputs benefit from a final verification pass.

A Real-World Prompt Using Multiple Commands

Here is what a well-structured, multi-command Claude prompt looks like in practice — for a developer writing a technical blog post:

// Step 1: Set the context /context I am a backend developer writing for a technical audience of senior engineers. The blog is focused on Go and distributed systems. Tone: direct, no fluff. // Step 2: Research the topic /research Latest best practices for distributed tracing in Go microservices, 2026. // Step 3: Extract the most important points /key-points What are the five most actionable insights from that research? // Step 4: Generate a draft /write A 900-word technical blog post based on those five points. Include one code example per section. Use H2 headers. End with a practical checklist. // Step 5: Refine /improve Make the introduction more direct. Cut the third paragraph by half.

Five commands. One high-quality, ready-to-publish technical blog post. This is the compounding effect of structured command use.

FAQ Section- “People Also Ask”

Q1 — Targets: “does Claude AI have commands”
Does Claude have built-in commands? Yes — Claude supports 89 structured commands organized across 11 categories including writing, code, analysis, research, and agentic workflows. Unlike ChatGPT’s phrase-based approach, Claude’s commands follow a consistent /command format that makes prompting faster and more predictable.
Q2 — Targets: “Claude vs ChatGPT which is better for coding”
Is Claude better than ChatGPT for coding? Claude’s 200K+ token context window makes it uniquely suited for large codebase analysis, refactoring, and documentation. For complex multi-file code understanding, Claude consistently outperforms in depth and accuracy.
Q3 — Targets: “what is Claude AI used for”
What is Claude AI best used for? Claude excels at long-form writing, complex code analysis, document summarization, agentic task execution via Claude Code CLI, and any task requiring large context — making it especially powerful for developers, researchers, and content creators.

Every Command Is a Shortcut to Clarity

The 89 commands documented here are not a list to memorize — they are a vocabulary to build. Start with five, master them, add five more. Within a month, you will have a personal command stack that makes every interaction with Claude faster, sharper, and more useful. The gap between a casual user and a power user of Claude is not intelligence — it is vocabulary.

Start Using Claude →
🤖
PrepCampusPlus AI Tutor
Scroll to Top