ScaleNix FOSS Teams

📚 Table of Contents

🎯 Introduction - Why interactive mode is the real power move
🚀 Launching Interactive Mode - Start, navigate, and exit
⚡ Essential Slash Commands - Your control panel
⌨️ Keyboard Shortcuts - Navigate faster
💬 Multi-Turn Strategies - Build context
📁 Working with Files - File operations
🎮 Real-World Workflow - Complete example
❓ FAQ - Common questions
🎓 Conclusion - Next steps

🎯 Introduction

Running claude once is cool—but the real magic happens in interactive mode. This is where Claude transforms from a simple question-answering tool into your AI pair programmer 🤖, maintaining context across conversations, understanding your project deeply, and executing complex multi-step workflows.

What Makes Interactive Mode Different? 🌟

| One-Shot Mode | Interactive Mode | |---------------|------------------| | Single question → answer → done | Continuous conversation with memory | | No context between commands | Context builds with each exchange | | Start from scratch every time | Resume where you left off | | Good for quick tasks | Perfect for complex workflows |

In this Claude CLI interactive mode guide, you'll master the commands and workflows that separate casual users from power users. You'll learn slash commands that control your sessions, keyboard shortcuts that save hours, and conversation strategies that get better results from Claude.

By the end of this tutorial, you'll work with Claude CLI like a seasoned pro, building context strategically and solving problems faster than ever. 💪

Let's dive into interactive mode mastery.

🚀 Launching Interactive Mode

Interactive mode is your default workspace when using Claude CLI. It's where you'll spend most of your time.

Starting a Session

Navigate to your project directory and launch Claude:

cd ~/projects/my-app
claude

You'll see the Claude Code welcome screen:

╭──────────────────────────────────────────────────────────╮
│                                                          │
│   ██████╗██╗      █████╗ ██╗   ██╗██████╗ ███████╗     │
│  ██╔════╝██║     ██╔══██╗██║   ██║██╔══██╗██╔════╝     │
│  ██║     ██║     ███████║██║   ██║██║  ██║█████╗       │
│  ██║     ██║     ██╔══██║██║   ██║██║  ██║██╔══╝       │
│  ╚██████╗███████╗██║  ██║╚██████╔╝██████╔╝███████╗     │
│   ╚═════╝╚══════╝╚═╝  ╚═╝ ╚═════╝ ╚═════╝ ╚══════╝     │
│                                                          │
╰──────────────────────────────────────────────────────────╯

Model: Claude Sonnet 4.5
Project: my-app
Session: Started Nov 4, 2025

Recent Conversations:
  1. Fixed authentication bug (2 hours ago)
  2. Added user profile feature (Yesterday)

Type /help for commands or start chatting...

>

Understanding the Interface

Your interactive session has three key areas:

📊 Status Line (top):

Current model (Claude Sonnet 4.5)
Project name (from your directory)
Session start time

📝 Recent Conversations:

Quick access to previous sessions
Resume with /resume command
Automatic conversation history

⌨️ Prompt (>):

Type questions, commands, or requests
Multi-line input supported (Shift+Enter)
Paste images directly (Ctrl+V)

Exiting Interactive Mode (Three Clean Ways)

| Method | Keyboard | When to Use | |--------|----------|-------------| | 🟢 Recommended | Ctrl+D at empty prompt | Normal exit | | 🔵 Command | /exit | Explicit shutdown | | 🔴 Emergency | Ctrl+C twice | Cancel operation first, then exit |

Pro Tip 💡: Always use Ctrl+D or /exit for clean shutdowns. This saves your conversation history properly.

⚡ Essential Slash Commands

Slash commands are your control panel for interactive sessions. Think of them as keyboard shortcuts for common tasks. Type /help to see all available commands.

Command Categories (Three Groups)

Commands fall into three natural groups:

🎮 Session Control: /clear, /exit, /resume - Manage your workflow
📊 Status & Info: /help, /status, /cost, /doctor - Check system health
🛠️ Advanced Tools: /compact, /init, /memory, /config - Power user features

🎮 Session Management Commands

`/clear` - Reset Conversation History

Removes all context from the current session. Use when:

| Situation | Why Clear Helps | |-----------|----------------| | Starting completely new task | No confusing old context | | Claude seems confused | Fresh mental model | | Hit token limits | Free up memory | | After long debugging session | Start with clean slate |

> /clear
✓ Conversation history cleared
Ready for a fresh start!

>

Warning ⚠️: You lose all conversation context. Claude won't remember anything from before /clear.

`/exit` - End Session

Cleanly exits interactive mode and saves your conversation.

> /exit
Session saved. See you next time!

`/resume` - Continue Previous Conversation

Resume a recent conversation with full context restored.

> /resume
Recent conversations:
  1. Fixed authentication bug (2 hours ago)
  2. Added user profile feature (Yesterday)
  3. Refactored database queries (2 days ago)

Select conversation (1-3): 1

✓ Resumed: Fixed authentication bug
Context restored from 2 hours ago.

>

Use cases:

Continuing work from yesterday
Getting back to an unfinished task
Reviewing what you did previously

📊 Status & Information Commands

`/help` - Display Available Commands

Shows all slash commands, including custom commands from your .claude/commands/ directory.

> /help

Built-in Commands:
  /clear      - Reset conversation history
  /exit       - End current session
  /help       - Show this help message
  /config     - Open configuration
  /status     - Show account and system info
  /cost       - Display token usage
  /doctor     - Diagnose installation issues
  /compact    - Summarize conversation to extend context
  /resume     - Continue previous conversation
  /init       - Generate CLAUDE.md project guide
  /memory     - Edit project memory
  /mcp        - View MCP server status

Custom Commands:
  /review     - Perform code review
  /test       - Generate unit tests
  /docs       - Create documentation

Type /help <command> for detailed help on specific commands.

`/status` - Account & System Information

View your account details, current model, and system status.

> /status

Account: user@example.com (Claude Pro)
Model: Claude Sonnet 4.5
Project: my-app (/Users/dev/projects/my-app)
Session Duration: 47 minutes
Token Usage: 12,450 / 200,000 (6%)
MCP Servers: 2 connected

Useful for:

Checking token usage before long operations
Verifying which account you're using
Confirming project directory

`/cost` - Token Usage Statistics

See detailed breakdown of your token consumption.

> /cost

Session Token Usage:
─────────────────────────────────────
Input Tokens:     8,234
Output Tokens:    4,216
Total Tokens:    12,450
────────────────────────────────────
Estimated Cost:  $0.18
Remaining Quota: 187,550 tokens (94%)

Token Management Strategy 📊:

| Usage Level | Action to Take | |-------------|---------------| | 0-50% | Work freely | | 50-80% | Start planning to compact | | 80-95% | Use /compact before next big task | | 95%+ | Consider /clear or new session |

Pro Tip 💡: Check /cost before complex tasks. If you're near your limit, consider using /compact to summarize the conversation and free up context.

🛠️ Advanced Workflow Commands

`/compact` - Summarize Conversation

When you're hitting context limits but want to continue, /compact has Claude summarize the conversation, preserving key information while freeing up tokens.

> /compact

Summarizing conversation...

✓ Conversation compacted
Previous context: 45,000 tokens
New context: 8,500 tokens
Saved: 36,500 tokens (81% reduction)

Summary preserved:
- Authentication bug investigation
- Fixed JWT expiration issue
- Updated tests
- Ready to continue

>

When to use /compact:

Long debugging sessions
After completing a major task
Before starting something new without losing context

`/init` - Generate CLAUDE.md Project Guide

Creates a CLAUDE.md file that provides context about your project to future Claude sessions.

> /init

Analyzing your project...
✓ Read 47 files
✓ Identified architecture patterns
✓ Detected tech stack: React, Node.js, PostgreSQL

Creating CLAUDE.md...
✓ Project overview added
✓ Architecture documented
✓ Common commands listed

CLAUDE.md created! Future sessions will use this guide.

What gets included:

Project overview and purpose
Tech stack and dependencies
Architecture patterns
Common development commands
Testing strategies

`/memory` - Edit Project Memory

Customize what Claude remembers about your project across sessions.

> /memory

Opening project memory...

Current memory:
- Uses React 18 with TypeScript
- API endpoints in /api directory
- Testing with Jest
- Deployment via Docker

Add custom notes or edit existing memory.
Press Ctrl+S to save, Ctrl+C to cancel.

Use cases:

Document team conventions
Remember specific quirks
Store frequently needed context

`/config` - Configuration Interface

Opens interactive configuration for Claude CLI settings.

> /config

Claude Code Configuration
─────────────────────────────────────
Model Settings:
  [x] Auto-select model
  [ ] Always use Sonnet 4.5
  [ ] Always use Opus

Editor Settings:
  Editor: vim
  [ ] Auto-accept suggestions

Display Settings:
  [x] Show status line
  [x] Syntax highlighting
  [ ] Verbose mode

Press Enter to select, 's' to save, 'q' to quit

`/doctor` - Diagnose Issues

Runs diagnostic checks on your Claude CLI installation.

> /doctor

Running diagnostics...

✓ Claude CLI version: 2.1.0 (latest)
✓ Authentication: Active
✓ API connectivity: Healthy
✓ Project detection: Working
⚠ Git integration: No .git directory found
✓ File permissions: OK

1 warning found. Would you like details? (y/n)

Use /doctor when:

Commands aren't working as expected
Authentication issues arise
File access problems occur

🎨 Custom Slash Commands

Create your own commands by adding .md files to .claude/commands/:

Example: .claude/commands/review.md

---
description: Perform comprehensive code review
---
Review this code for:
- Bugs and edge cases
- Performance issues
- Security vulnerabilities
- Best practices
- Code style consistency

Provide specific suggestions with code examples.

Usage:

> /review @src/auth.js

Running code review on src/auth.js...

## Review: src/auth.js

### Security Issues (High Priority)
Line 23: JWT secret is hardcoded...

Learn more about custom commands in our Custom Commands tutorial.

📋 Quick Command Reference

| Command | One-Line Purpose | Common Scenario | |---------|-----------------|-----------------| | /clear | Wipe conversation history | Switching tasks | | /exit | End session | Done for the day | | /resume | Continue old conversation | Pick up where you left off | | /help | Show all commands | When you forget syntax | | /status | Check account & tokens | Before big tasks | | /cost | See token usage | Monitor spending | | /compact | Compress context | Nearing token limit | | /init | Create project guide | New project setup | | /memory | Edit project notes | Store team conventions | | /config | Change settings | Customize behavior | | /doctor | Diagnose problems | Troubleshooting |

Mind map showing slash commands organized by category: Session commands, Status commands, Project commands, and Custom commands Figure 2: Complete slash commands reference map showing all built-in and custom command categories available in Claude CLI interactive mode.

⌨️ Keyboard Shortcuts for Power Users

Master these shortcuts to navigate Claude CLI faster than ever. Think of them as hotkeys that save you seconds hundreds of times per day. 🚀

Essential Navigation Shortcuts

| Shortcut | Action | Use Case | |----------|--------|----------| | Ctrl+R | Reverse search history | Find previous commands | | Ctrl+D | Exit Claude Code | Clean shutdown | | Esc | Interrupt Claude | Cancel long operations | | Esc Esc | Rewind/Browse history | Restore earlier state | | Ctrl+V | Paste image | Add screenshots | | Ctrl+B | Run in background | Continue work while commands run | | Shift+Tab | Auto-accept | Accept suggestion immediately | | ↑/↓ | Navigate history | Browse previous inputs |

Ctrl+R: Reverse Search (Your New Best Friend)

Press Ctrl+R and start typing to search your command history:

(reverse-i-search): auth
> Explain how authentication works in src/auth.js

How it works:

Press Ctrl+R
Type search term (e.g., "auth")
Press Ctrl+R again to cycle through matches
Press Enter to select
Edit if needed, then execute

Pro Tip 💡: Use descriptive prompts so they're easy to find later. Instead of "fix this", use "fix authentication token expiration bug".

Esc Esc: Time Travel for Your Code

Double-tap Esc at an empty prompt to browse conversation history and restore to an earlier point:

> (empty prompt)
Esc Esc

Conversation History:
────────────────────────────────────
  5. Updated tests for auth module
  4. Fixed JWT expiration bug
  3. Analyzed authentication flow
  2. Reviewed src/auth.js
  1. Explained project structure

Select checkpoint (1-5): 4

✓ Restored to: Fixed JWT expiration bug
Code state and conversation context restored.

When to use:

You tried something that didn't work
Want to explore a different approach
Made changes you want to undo

Ctrl+V: Paste Images for Visual Context

Take a screenshot and paste it directly into Claude:

# macOS: Cmd+Ctrl+Shift+4 (screenshot to clipboard)
# Then: Ctrl+V in Claude CLI

> (Ctrl+V to paste screenshot)
📷 Image pasted

> This error dialog is showing. What's wrong and how do I fix it?

Analyzing screenshot...

I can see a React error: "Cannot read property 'map' of undefined"
This happens when your data array isn't loaded yet...

Use cases:

UI bugs with visual elements
Error dialogs
Design mockups
Database schema diagrams

Ctrl+B: Background Command Execution

Run commands in the background while continuing to work with Claude:

> Let me start the dev server in the background
Ctrl+B npm run dev

✓ Command running in background (shell_1)
Output available in /bashes

> Now let's continue building the API while that runs...

Common background tasks:

Development servers (npm run dev)
Log monitoring (tail -f app.log)
Long builds (npm run build)
Test watchers (npm test -- --watch)

Bash-Style Input Navigation

Claude CLI respects standard bash shortcuts:

| Shortcut | Action | |----------|--------| | Ctrl+A | Jump to beginning of line | | Ctrl+E | Jump to end of line | | Ctrl+W | Delete word backwards | | Ctrl+U | Delete entire line | | Opt+F | Jump forward one word (macOS) | | Opt+B | Jump backward one word (macOS) | | Alt+F | Jump forward one word (Linux) | | Alt+B | Jump backward one word (Linux) |

Keyboard Shortcuts Quick Reference

Three groups of shortcuts:

🔍 Search & History: Ctrl+R, ↑/↓, Esc Esc - Find past work
⚙️ Control Flow: Ctrl+D, Esc, Ctrl+B - Manage execution
✏️ Text Editing: Ctrl+A/E/W/U, Opt/Alt+F/B - Edit prompts

💬 Multi-Turn Conversation Strategies

Interactive mode shines when you build context strategically across multiple turns. Think of it like pair programming—the more Claude understands your project, the better the solutions. 🤝

Strategy 1: Progressive Context Building

Start broad, then narrow down:

Turn 1: > What does this project do?
Turn 2: > How is authentication handled?
Turn 3: > The JWT tokens are expiring too quickly. Why?
Turn 4: > Fix the expiration time to 7 days

Pattern: Project overview → Subsystem → Specific issue → Precise fix

Why this works: Claude builds a mental model of your project, with each turn adding specificity for more accurate final requests.

Strategy 2: Reference Previous Responses

Claude remembers the conversation. Reference earlier exchanges:

> Analyze the authentication flow
> Can you write tests for the flow you just described?

Key phrase patterns:

"Based on what you found earlier..."
"Like you explained before..."
"Following the same pattern you showed..."
"Using the approach from your previous suggestion..."

Strategy 3: Iterative Refinement

Don't expect perfection on the first try. Iterate in layers:

Turn 1: > Create a user profile component
Turn 2: > Add form validation
Turn 3: > Make the error messages more user-friendly
Turn 4: > Add loading states during form submission

Pattern: Feature → Enhancement → Refinement → Polish

Strategy 4: When to Start Fresh

Sometimes you need to clear context and start over:

Clear context when:

Switching to completely unrelated task
Claude seems confused by old conversation
You've gone down the wrong path
After solving a complex problem
Nearing token limits

> /clear
> Let's start fresh. I need to build a new feature for...

Token Management Strategies

Long conversations consume tokens. Manage them wisely:

Check usage periodically:

> /cost

Compact when nearing limits:

> /compact

Clear after completing tasks:

> /clear

Strategic summarization:

> Summarize what we've accomplished so far
> /clear
> Here's the context: [paste summary]
Now let's continue with...

Conversation Strategy Quick Reference

Five key strategies:

📊 Progressive Context - Complex investigations, builds deep understanding
🔗 Reference Previous - Iterating on solutions, maintains continuity
🔄 Iterative Refinement - Creating features, layers improvements
🧹 Start Fresh - Switching tasks, clears confusion
💰 Token Management - Long sessions, avoids limits

📁 Working with Files in Interactive Mode

Claude CLI understands your project structure and can work with files intelligently. No more copying and pasting between terminal and editor! 🎉

Reading Files (Three Approaches)

| Method | Syntax | When to Use | |--------|--------|-------------| | 🤖 Automatic | Ask naturally | Let Claude find relevant files | | 🎯 @-mention | @src/auth.js | Explicitly specify file | | 📦 Multiple | @file1 @file2 | Compare or analyze together |

Automatic reading: Claude automatically reads relevant files when you ask questions:

> Explain what src/auth.js does

Claude reads src/auth.js and explains...

Explicit @-mentions: Reference files directly for precise control:

> Review @src/auth.js for security issues

Claude focuses specifically on that file...

Multiple files:

> Compare @src/auth.js and @src/api/auth.ts

Claude analyzes both files...

Editing Files Interactively (The Review Workflow)

Ask Claude to modify code, then review before applying:

> Add error handling to src/auth.js

Claude proposes changes:

Would you like me to apply these changes to src/auth.js?
(y/n/edit):

Your three options:

| Option | What Happens | When to Use | |--------|--------------|-------------| | ✅ y | Apply immediately | Changes look good | | ❌ n | Reject changes | Not what you wanted | | 📝 edit | Review in editor first | Want to see full context |

Creating New Files

> Create a utility function file at src/utils/dates.js
  with functions to format dates

Claude generates the file and writes it:

✓ Created src/utils/dates.js
Added functions:
  - formatDate(date, format)
  - parseDate(string)
  - isValidDate(date)

Would you like me to write tests as well?

File Path Handling (All Forms Work)

Claude understands multiple path formats:

| Path Type | Examples | When to Use | |-----------|----------|-------------| | 📍 Absolute | /Users/dev/projects/my-app/src/auth.js | Unambiguous | | 📂 Relative | src/auth.js or ./src/auth.js | From project root | | 🎯 @-mention | @src/auth.js | Explicit reference | | 🌐 Glob | src/**/*.test.js | Multiple files |

Glob patterns for bulk operations:

> Review all TypeScript files in src/api
> Analyze @src/**/*.test.js

File Operations Quick Reference

Five common operations:

Read single file: Explain @src/auth.js
Read multiple: Compare @src/auth.js @src/api/auth.ts
Edit existing: Add error handling to src/auth.js
Create new: Create utility file at src/utils/dates.js
Bulk analysis: Review all test files

🎮 Real-World Workflow Example

Let's walk through a complete debugging session to see these concepts in action. This is how you'll actually use Claude CLI in daily work. 🔧

Scenario: React Component Re-rendering Bug

You have a performance issue in your app. Let's solve it with Claude CLI using proper multi-turn strategy.

Turn 1: Start the Investigation 🔍

> I have a React component that's re-rendering too often.
  Let me show you the component file.

Claude: I'd be happy to help! Please share the component file, and I'll analyze the rendering behavior.

> @src/components/UserProfile.js

Claude reads the file and responds:

Looking at UserProfile.js, I can see several potential causes:

1. Line 15: useEffect with no dependency array runs on every render
2. Line 23: Inline arrow function in onClick creates new function each render
3. Line 31: Object literal in style prop creates new object each render

These are causing unnecessary re-renders. Would you like me to fix them?

What happened: Progressive context building—Claude now understands the problem deeply.

Turn 2: Apply Fixes ⚙️

> Yes, please fix all three issues

Claude: I'll optimize the component to prevent unnecessary re-renders.

✓ Applied changes to src/components/UserProfile.js

Changes made:
1. Added dependency array to useEffect
2. Wrapped click handler with useCallback
3. Moved style object outside component
4. Added React.memo for props comparison

Would you like me to show you the diff?

> Yes

Claude shows before/after diff...

> Looks good. Can you also add comments explaining the optimizations?

Claude: Adding explanatory comments...

✓ Updated src/components/UserProfile.js with comments

What happened: Iterative refinement—started with fix, then enhanced with comments.

Turn 3: Verify the Fix ✅

> Now help me verify this actually fixed the performance issue.
  What should I look for?

Claude: Here's how to verify:

1. Open React DevTools Profiler
2. Record a session while interacting with UserProfile
3. Look for reduced render count

Let me also create a simple performance test:

✓ Created src/components/UserProfile.test.js

This test verifies:
- Component only re-renders when props actually change
- Callbacks maintain stable references
- Style objects don't cause re-renders

Run: npm test -- UserProfile.test.js

What happened: Reference previous work—Claude builds on the fixes to create tests.

Turn 4: Run Tests and Document 📝

Ctrl+B npm test -- UserProfile.test.js

✓ Command running in background

> While tests are running, can you document these optimizations
  in the component file as JSDoc comments?

Claude: Adding comprehensive JSDoc...

✓ Updated with JSDoc comments explaining:
  - Performance considerations
  - When re-renders occur
  - Memoization strategies

> /cost

Token Usage: 15,420 tokens
Remaining: 184,580 (92%)

> Perfect. Let me check the test results.

Claude: The tests passed! Your component is now optimized.

✓ All tests passed (3/3)
  - Prevents unnecessary re-renders
  - Maintains stable callback references
  - Style objects properly memoized

Would you like me to apply similar optimizations to other components?

What happened: Background execution + token management—worked efficiently.

Session Summary: In 10 minutes, you:

✅ Identified 3 performance issues
✅ Applied fixes with explanations
✅ Created performance tests
✅ Added documentation
✅ Verified the solution

All without leaving your terminal.

Interactive workflow showing multi-turn conversation between developer, Claude CLI, and Claude AI with context building across turns Figure 1: Interactive mode workflow showing how context builds across multiple conversation turns, enabling Claude to understand your project deeply and deliver increasingly accurate solutions.

❓ FAQ

How do I see my conversation history?

When you start claude, recent conversations appear on the welcome screen. Use /resume to continue any previous conversation with full context restored.

Can I save and resume conversations?

Yes! Claude CLI automatically saves conversations. Use /resume to see recent sessions and continue where you left off. Conversations persist across terminal sessions.

What happens when I hit token limits?

Claude CLI will warn you when approaching limits. Options:

Use /compact to summarize and compress context
Use /clear to start fresh
Start a new session with a summary of previous work

Check current usage anytime with /cost.

How do I clear context and start fresh?

Use the /clear command to remove all conversation history from the current session. Claude will have no memory of previous exchanges, giving you a clean slate for new tasks.

Can I use Claude CLI with multiple projects simultaneously?

Yes, but each terminal window is one session. Open multiple terminals for multiple projects:

# Terminal 1
cd ~/projects/frontend
claude

# Terminal 2
cd ~/projects/backend
claude

Each maintains separate context and conversation history.

How do keyboard shortcuts work on Windows vs macOS?

Most shortcuts are the same. Key differences:

macOS: Use Opt+F/B for word jumping
Windows/Linux: Use Alt+F/B for word jumping

What's the difference between /clear and /exit?

/clear - Removes conversation history but keeps session active. You stay in Claude CLI.
/exit - Ends the session entirely and returns to your shell.

Use /clear when switching tasks. Use /exit when you're done working.

🎓 Conclusion

You've just leveled up from casual Claude CLI user to interactive mode power user. 🚀 You now know:

What You've Mastered

✅ Essential slash commands for session control
✅ Keyboard shortcuts that save hours
✅ Multi-turn conversation strategies for better results
✅ File operations and project navigation
✅ Real debugging workflows that deliver results

Interactive mode isn't just about asking questions—it's about building a working relationship with Claude where context accumulates and solutions emerge through conversation.

🎯 What's Next?

In the next post, "Automate Your Workflow: Claude CLI Bash Scripting Guide", you'll learn:

Print mode (-p) for scriptable output
Piping data to/from Claude
Creating automation scripts
Git workflow automation
Real-world automation examples

Interactive mode is powerful. But when you combine it with automation? That's when Claude CLI becomes a productivity multiplier. 📈

Practice Challenge

Before moving on, try this 15-minute exercise:

Start claude in one of your projects
Use /status to check your setup
Ask: "What does this project do?"
Ask a follow-up question referencing Claude's response
Try Ctrl+R to search your conversation
Check /cost to see token usage
Use /compact to compress the conversation
Continue with a new question

Notice how context builds and Claude's responses become more precise. That's the power of interactive mode.

📚 Resources

Previous Post: Claude CLI Tutorial - Getting Started
Official Interactive Mode Docs
Slash Commands Reference
Community Workflows

Next in Series: Automate Your Workflow: Claude CLI Bash Scripting Guide →

Found this helpful? Share it with your team and let's master Claude CLI together.

Questions about interactive mode? Drop a comment below—I respond to every one.

Last Updated: November 2025 | Claude Code Version 2.x | Part 2 of 8 in the Claude CLI Zero to Hero Series

📚 Table of Contents

🎯 Introduction

What Makes Interactive Mode Different? 🌟

🚀 Launching Interactive Mode

Starting a Session

Understanding the Interface

Exiting Interactive Mode (Three Clean Ways)

⚡ Essential Slash Commands

Command Categories (Three Groups)

🎮 Session Management Commands

/clear - Reset Conversation History

/exit - End Session

/resume - Continue Previous Conversation

📊 Status & Information Commands

/help - Display Available Commands

/status - Account & System Information

/cost - Token Usage Statistics

🛠️ Advanced Workflow Commands

/compact - Summarize Conversation

/init - Generate CLAUDE.md Project Guide

/memory - Edit Project Memory

/config - Configuration Interface

/doctor - Diagnose Issues

🎨 Custom Slash Commands

📋 Quick Command Reference

⌨️ Keyboard Shortcuts for Power Users

Essential Navigation Shortcuts

Ctrl+R: Reverse Search (Your New Best Friend)

Esc Esc: Time Travel for Your Code

Ctrl+V: Paste Images for Visual Context

Ctrl+B: Background Command Execution

Bash-Style Input Navigation

Keyboard Shortcuts Quick Reference

💬 Multi-Turn Conversation Strategies

Strategy 1: Progressive Context Building

Strategy 2: Reference Previous Responses

Strategy 3: Iterative Refinement

Strategy 4: When to Start Fresh

Token Management Strategies

Conversation Strategy Quick Reference

📁 Working with Files in Interactive Mode

Reading Files (Three Approaches)

Editing Files Interactively (The Review Workflow)

Creating New Files

File Path Handling (All Forms Work)

File Operations Quick Reference

🎮 Real-World Workflow Example

Scenario: React Component Re-rendering Bug

Turn 1: Start the Investigation 🔍

Turn 2: Apply Fixes ⚙️

Turn 3: Verify the Fix ✅

Turn 4: Run Tests and Document 📝

❓ FAQ

How do I see my conversation history?

Can I save and resume conversations?

What happens when I hit token limits?

How do I clear context and start fresh?

Can I use Claude CLI with multiple projects simultaneously?

How do keyboard shortcuts work on Windows vs macOS?

What's the difference between /clear and /exit?

🎓 Conclusion

What You've Mastered

🎯 What's Next?

Practice Challenge

📚 Resources

`/clear` - Reset Conversation History

`/exit` - End Session

`/resume` - Continue Previous Conversation

`/help` - Display Available Commands

`/status` - Account & System Information

`/cost` - Token Usage Statistics

`/compact` - Summarize Conversation

`/init` - Generate CLAUDE.md Project Guide

`/memory` - Edit Project Memory

`/config` - Configuration Interface

`/doctor` - Diagnose Issues