CLAUDE MASTERY PLAYBOOK

Chat. Cowork. Code. The complete guide to every mode, feature, shortcut, and expert workflow — plus what your Max plan actually gives you and how it powers Boots On The Ground AI.

18 Sections
10 Golden Rules
Skill Grading
Feb 2026
Table of Contents
1

The Claude Ecosystem

Claude isn't just one thing. It's three distinct tools, each designed for a different way of working. Think of them as three gears — same engine, different speeds.

┌──────────────────────────────────────────────────────────┐ │ CLAUDE DESKTOP APP │ │ │ │ ┌─────────────┐ ┌──────────────┐ ┌──────────────┐ │ │ │ 💬 CHAT │ │ 🤝 COWORK │ │ ⌨ CODE │ │ │ │ │ │ │ │ │ │ │ │ Conversation│ │ Desktop Agent│ │ Terminal │ │ │ │ Q&A, Ideas │ │ File Actions │ │ Full Dev │ │ │ │ Research │ │ Automation │ │ Environment │ │ │ │ │ │ │ │ │ │ │ │ Free/Pro/Max│ │ Max Only │ │ Pro/Max │ │ │ └─────────────┘ └──────────────┘ └──────────────┘ │ │ │ │ Lowest Tokens ──────────────────── Highest Tokens │ │ Least Power ──────────────────── Most Power │ └──────────────────────────────────────────────────────────┘
⚡ Key Insight

Each mode consumes tokens differently. Chat is the lightest. Cowork uses more because it plans and executes multi-step tasks. Code uses the most because it reads files, writes code, runs commands, and orchestrates agents — all in tokens.

2

Claude Chat — The Foundation

This is the Claude most people know. Open claude.ai or the desktop app, type a question, get an answer.

What Chat does well

What Chat cannot do

Chat features worth knowing

FeatureWhat It DoesTier
ProjectsOrganize conversations with shared context docsPro+
ArtifactsInteractive code previews, documents, diagramsAll
Web SearchReal-time internet searches in conversationAll
File UploadAnalyze PDFs, images, CSVs, code filesAll
Extended ThinkingDeeper reasoning for complex problemsPro+
MemoryRemembers preferences across conversationsMax
Google WorkspaceConnect to Google Docs, Drive, CalendarPro+
✅ Pro Tip

Use Projects to give Claude persistent context. Upload your brand guidelines, product docs, or code standards into a project, and every conversation in that project has that context automatically — without you re-explaining every time.

3

Claude Cowork — The Desktop Agent

Cowork is Claude Code's consumer-friendly sibling. It runs inside the Claude desktop app and can actually do things on your computer — create files, organize folders, process documents — all through a visual interface.

What makes Cowork different

ChatCowork
Answers questionsCompletes tasks
Suggests what to doActually does it
You copy-paste resultsClaude writes files directly
One response at a timePlans multi-step workflows
Lightweight usageHeavier token usage

Great use cases for Cowork

⚠️ Important Limitations

Max plan required ($100-$200/month). macOS only (for now). No memory across sessions. The desktop app must stay open while tasks run. Cowork uses significantly more tokens than Chat because of multi-step task execution.

⚡ Boots On The Ground AI Workflow

This is how Damara creates playbooks — Cowork handles the branding, formatting, and content generation, then outputs the files that get pushed to the GitHub Pages repo. Chat would just tell you what to do. Cowork actually does it.

4

Claude Code — The Terminal Powerhouse

Claude Code is the most powerful way to use Claude. It runs in your terminal and has full access to your development environment — it can read your codebase, write files, run commands, execute tests, commit to git, deploy to production, and orchestrate teams of AI agents.

What Claude Code can do that nothing else can

What happens when you type in Claude Code: You: "Add a search bar to the dashboard" Claude Code: 1. 🔍 Reads your codebase (finds components, styles, patterns) 2. 🛠 Plans the approach (which files to modify) 3. ✏ Writes the code (creates/edits files) 4. ▶ Runs it (starts dev server, checks for errors) 5. 🔧 Fixes issues (reads errors, adjusts code) 6. ✅ Verifies (runs tests, confirms it works) All in one conversation. All from your terminal.
💡 Why Terminal?

The terminal gives Claude Code direct access to your system. No browser sandbox, no file upload limits, no copy-paste. It works WITH your tools — your git, your package manager, your deployment pipeline. That's why it's more powerful than Chat or Cowork for development work.

5

Getting Started with Claude Code

Installation

# Install Claude Code globally npm install -g @anthropic-ai/claude-code # Verify installation claude --version # Check for issues claude /doctor # Launch Claude Code claude # Launch in a specific project cd ~/my-project && claude

First-time setup checklist

  1. Install Node.js (v18+) if you don't have it
  2. Install Claude Code via npm
  3. Run claude and authenticate with your Anthropic account
  4. Navigate to a project folder and run claude from there
  5. Create a CLAUDE.md in your project root (Section 10)
✅ Pro Tip

Always launch Claude Code from your project's root directory. Claude reads the directory structure to understand your project. Launching from ~ (home) means it sees everything — which is slower and less focused.

6

Every Slash Command Explained

Slash commands are Claude Code's power menu. Type / to see them all.

Essential commands (use these daily)

CommandWhat It DoesWhen to Use It
/helpShows all available commandsWhen you forget a command
/compactCompresses conversation historyWhen context gets long, saves tokens
/costShows token usage and cost for sessionCheck how much you're spending
/modelSwitch models (sonnet, opus, haiku)Use haiku for simple tasks, opus for complex
/clearClear conversation historyStart fresh on a new task
/memoryEdit your CLAUDE.md fileUpdate project instructions

Power user commands

CommandWhat It DoesWhen to Use It
/reviewCode review of recent changesBefore committing, get a second opinion
/planEnter Plan Mode (read-only exploration)Research before making changes
/vimToggle Vim keybindingsIf you're a Vim user
/permissionsManage file access permissionsControl what Claude can read/write
/contextView current context usageSee how full your context window is
/doctorDiagnose installation issuesWhen something isn't working right
/agentsManage custom subagentsSet up specialized AI assistants
/rewindRoll back conversation and codeUndo mistakes, try a different approach
/statusView current session statusCheck model, mode, and settings
⚡ The /compact Secret

/compact is your best friend for long sessions. When your conversation gets long, context fills up and Claude starts losing earlier details. Running /compact compresses the history, freeing up space and reducing token costs on subsequent messages. Use it every 15-20 messages.

7

Keyboard Shortcuts & Vim Mode

Essential keyboard shortcuts

ShortcutAction
EnterSend message
Shift + EnterNew line (without sending)
Ctrl + CCancel current generation
Ctrl + DExit Claude Code
Ctrl + LClear terminal screen
Esc + EscEdit your previous message
Shift + TabCycle through permission modes
Ctrl + TToggle task list

Vim Mode

If you're a Vim user, type /vim to enable Vim keybindings in the input area. This gives you:

Custom keybindings

# Open keybindings config /keybindings # This creates/opens ~/.claude/keybindings.json # Changes apply immediately — no restart needed
✅ Pro Tip

Esc + Esc (double escape) to edit your last message is a game-changer. Made a typo? Want to rephrase? Double-tap Esc instead of retyping the whole thing.

8

Permission Modes

Claude Code has five permission modes that control how much freedom Claude has. Think of them as safety levels.

ModeWhat Claude Can DoBest For
DefaultAsks permission for every file edit and commandLearning, unfamiliar codebases
Accept EditsAuto-approves file edits, asks for commandsDay-to-day development
Plan ModeRead-only — can search and read, cannot change anythingExploring before building
Don't AskAuto-approves everythingTrusted automation tasks
BypassNo restrictions at allIsolated environments only (containers)
# Cycle through modes Press Shift + Tab to cycle: Default → Accept Edits → Plan → ... # Or set via slash command /permissions
🚨 Safety Warning

Never use "Bypass" mode unless you're in an isolated container or VM. It gives Claude unrestricted access to run any command — including destructive ones. "Don't Ask" is nearly as risky. Stick with Default or Accept Edits for normal work.

9

Plan Mode — Think Before You Build

Plan Mode is one of Claude Code's most underrated features. It forces Claude to research and think before making any changes.

How it works

Normal Mode: Plan Mode: You ask ──> Claude You ask ──> Claude acts immediately reads, searches, plans presents approach YOU approve THEN Claude acts

What Claude can do in Plan Mode

What Claude CANNOT do in Plan Mode

⚡ Expert Pattern

The best developers use Plan Mode like a Senior Architect review. Start every significant task in Plan Mode: let Claude explore, understand, and propose. Review the plan. THEN switch to normal mode to execute. This prevents wasted work and token costs from going down the wrong path.

10

CLAUDE.md — Your Project Brain

CLAUDE.md is a special file that Claude Code reads automatically every time you start a session in that project. It's your project's instruction manual for Claude.

Where to put it

# Project-level (most common) ~/my-project/CLAUDE.md # Claude reads this when you run `claude` in my-project/ # Global (applies to ALL projects) ~/.claude/CLAUDE.md # Always loaded, every session # Nested (for specific directories) ~/my-project/src/CLAUDE.md # Extra context for the src/ directory

What to put in CLAUDE.md

# Example CLAUDE.md for a project # Project: Boots On The Ground AI Website ## Tech Stack - Next.js 14 with App Router - TypeScript (strict mode) - Tailwind CSS - Supabase for database - Deployed on Fly.io ## Coding Standards - Use TypeScript for all new files - Components go in src/components/ - Use server components by default - Client components only when needed (interactivity) ## Commands - Dev server: npm run dev - Build: npm run build - Test: npm test - Deploy: fly deploy ## Important Notes - Never modify .env files - Always run tests before committing - Use conventional commit messages - Brand colors: navy #1B2A4A, gold #D4A843
✅ Pro Tip

Think of CLAUDE.md like onboarding a new developer. What would they need to know on day one? Put that in CLAUDE.md. But keep it concise — CLAUDE.md is loaded as context every session, which means every token in it costs you on every message (see Token Management Playbook).

11

Custom Commands & Skills

Custom slash commands

You can create your own slash commands by adding Markdown files to .claude/commands/ in your project.

# Create a custom /deploy command mkdir -p .claude/commands cat > .claude/commands/deploy.md << 'EOF' Run the deployment checklist: 1. Run all tests 2. Build the project 3. Check for TypeScript errors 4. If everything passes, deploy to Fly.io 5. Verify the deployment is live EOF # Now you can type /deploy in Claude Code!

Skills (model-invoked)

Skills are more advanced than commands. They live in .claude/skills/ and Claude can invoke them automatically when relevant.

# .claude/skills/code-review.md --- name: code-review description: Reviews code for quality and security tools: [Read, Grep, Glob] --- Review the changed files for: - Security vulnerabilities (OWASP Top 10) - Performance issues - Code style consistency - Missing error handling - Test coverage gaps

Commands vs Skills

FeatureCommandsSkills
Triggered byYou type /commandClaude invokes automatically
ComplexitySimple prompt in one fileCan include tool restrictions, agents
Location.claude/commands/.claude/skills/
Parameters$ARGUMENTS placeholderYAML frontmatter config
12

Subagents & Agent Orchestration

This is where Claude Code becomes a team, not just a tool. Subagents are specialized AI assistants that Claude can spawn to handle specific tasks — each running in its own context window.

Built-in subagents

AgentPurposeTools Available
ExploreFast codebase search and analysisRead-only (Glob, Grep, Read)
PlanResearch and context gathering for plansRead-only
General PurposeComplex multi-step tasksFull toolset

Custom subagents

# .claude/agents/security-reviewer.md --- name: security-reviewer description: Reviews code for security vulnerabilities tools: [Read, Grep, Glob] model: sonnet --- You are a security expert. Review code for: - SQL injection, XSS, CSRF - Hardcoded secrets or credentials - Insecure authentication patterns - Missing input validation - OWASP Top 10 vulnerabilities

Agent Teams (experimental)

The most advanced orchestration pattern. Claude creates a team lead, spawns independent teammates, and coordinates their work through a shared task list.

Agent Teams Workflow: Team Lead (Claude Code) ├── Teammate 1: "Build the API endpoint" ├── Teammate 2: "Write the frontend component" └── Teammate 3: "Add tests for both" Each teammate: - Runs in its own context window - Has its own set of tools - Can see the shared task list - Flags dependencies to avoid conflicts - Reports back to the team lead
🚨 Token Warning

Agent orchestration is powerful but token-expensive. Each subagent has its own context window, meaning you're paying for multiple parallel conversations. A 3-agent team can easily use 3-5x the tokens of a single conversation. Use agents for genuinely parallelizable work, not simple sequential tasks.

✅ Best Practice

Limit yourself to 3-4 subagents maximum. More than that and the coordination overhead (also in tokens) outweighs the benefit. Give each agent a clear, specific task with defined scope.

13

MCP Servers — Supercharging Claude

MCP (Model Context Protocol) servers let Claude Code connect to external tools and services. Think of them as plugins that give Claude new superpowers.

What MCP servers enable

MCP ServerWhat Claude Can Do
Browser / PlaywrightNavigate websites, take screenshots, test UIs
Database (Postgres, etc.)Query databases, analyze data, run migrations
FigmaRead designs, implement UI from mockups
NotionRead/write docs, sync project info
GitHubManage issues, PRs, reviews
SentryAnalyze errors, debug production issues
SlackSend messages, read channels

Adding an MCP server

# Add an MCP server to your project claude mcp add # Example: Add a Postgres database connection claude mcp add postgres-server --connection-string $DATABASE_URL # Now Claude can query your database directly!
💡 Real-World Use

Connect Sentry via MCP, and you can say "Look at the top 5 errors in production and fix them." Claude reads the errors from Sentry, finds the code, and patches it. That's a workflow that used to take hours, done in minutes.

14

Expert Workflows

These are the patterns that separate casual users from power users.

1. The Plan-Execute-Review Loop

# Step 1: Enter Plan Mode /plan "Analyze the authentication system and propose improvements" # Step 2: Review the plan, then switch to normal mode Press Shift+Tab to switch back # Step 3: Execute "Implement the plan you just proposed" # Step 4: Review /review

2. Model Switching by Task

# Use Haiku for quick lookups (cheap, fast) /model haiku "What file handles the login route?" # Switch to Sonnet for implementation (balanced) /model sonnet "Add rate limiting to the login endpoint" # Switch to Opus for complex architecture (powerful) /model opus "Redesign the authentication system to support OAuth2 and SAML"

3. The Compact Ritual

# Every 15-20 messages, compress your context /compact # Check your costs periodically /cost # Start a new session for unrelated tasks /clear

4. CLI-First External Interactions

# Instead of describing what to do, tell Claude to use CLI tools "Use gh to create a pull request for my changes" "Use fly deploy to push this to production" "Use sentry-cli to check recent errors" "Use docker compose up to start the dev environment" # CLI tools are token-efficient — they return structured data # instead of Claude having to parse web pages

5. The /rewind Safety Net

# Made a wrong turn? Don't start over. /rewind # This rolls back both the conversation AND code changes # to an earlier checkpoint. Pick where to restore to. # Then try a different approach.
⚡ The Expert Mindset

Experts treat Claude Code like a senior developer on their team, not a magic box. They give clear context, review output critically, use Plan Mode to align before executing, switch models based on task complexity, and manage their token budget like a project budget.

15

Free vs Pro vs Max — The Full Breakdown

Feature Free Pro $20/mo Max $100-200/mo
Claude ChatYesYesYes
Claude CodeNoYesYes
Claude CoworkNoNoYes
Messages (per 5 hrs)~30-100~45 (standard)~225 (5x) / ~900 (20x)
Usage Multiplier1x1x5x or 20x vs Pro
Extended ThinkingNoYesYes
ProjectsLimitedUnlimitedUnlimited
MemoryNoNoYes
Priority AccessNoNoYes (peak times)
Early FeaturesNoNoYes
Google WorkspaceNoYesYes
Remote MCPNoYesYes
Web SearchYesYesYes
Extra Usage (API overflow)NoYes (pay as you go)Yes (pay as you go)

Max 5x ($100/mo) vs Max 20x ($200/mo)

FeatureMax 5xMax 20x
Messages per 5-hour window~225~900
Usage vs Pro5x more20x more
Claude Code usageGenerousVirtually unlimited
Cowork accessYesYes
Best forPower users, daily devsAll-day coding, agencies, heavy automation

Extra Usage (the overflow safety net)

When you hit your included limits, Claude doesn't stop — it switches to pay-as-you-go at standard API rates. You opt into this in settings. This means you never get cut off mid-task.

💡 The $50 Opus 4.6 Credit

Anthropic gave existing Pro and Max subscribers $50 in free Extra Usage credits to try Opus 4.6 (had to be claimed by Feb 16, 2026). This is the kind of early access perk Max subscribers regularly get.

16

What Max Gives Boots On The Ground AI

Here's specifically what the Max tier unlocks for Damara and the Boots On The Ground AI workflow:

The Boots playbook creation workflow

1. 💬 Chat — Brainstorm playbook topics and outline content 2. 🤝 Cowork — Generate branded playbook content, format files 3. ⌨ Code — Build the HTML, update index.html, push to GitHub Pages All three modes. All in one subscription. That's Max.

What Max enables that Pro doesn't

CapabilityHow Boots Uses It
Cowork (Max only)Create branded playbooks, process files, automate content generation without touching a terminal
5-20x usageBuild multiple playbooks in one day without hitting limits
MemoryClaude remembers Boots branding (navy #1B2A4A, gold #D4A843), coding preferences, project structure
Priority accessNo waiting during peak hours — critical when on a deadline
Early featuresFirst access to new Claude capabilities (Agent Teams, new models)
Extra UsageNever get cut off mid-playbook — overflow to API pricing seamlessly

The ROI math

// What Boots gets from Max ($100-200/month): Without Claude: - Hire a developer for playbooks: $2,000-5,000/playbook - Hire a designer for branding: $500-2,000/playbook - Manual GitHub Pages management: Hours of learning - Total for 7 playbooks: $17,500-$49,000 With Claude Max: - Monthly subscription: $100-200/month - 7 playbooks created in weeks: $200-400 total - Ongoing updates and new playbooks: Included - Learning full-stack development: Priceless
⚡ The Big Picture

Max isn't just "more messages." It's the difference between using Claude as a chatbot and using Claude as a full-stack development team. For Boots On The Ground AI, it means going from "I need to hire someone" to "I can build this myself" — for a fraction of the cost.

17

Grade Your Claude Skills

Where do you stand? Be honest. Use this as a roadmap — not a judgment. Every expert started at beginner.

Level 1: Beginner

Beginner Use Claude Chat to ask questions and get answers
Beginner Upload files for Claude to analyze
Beginner Know the difference between Chat, Cowork, and Code
Beginner Have Claude Code installed and can launch it

Level 2: Intermediate

Intermediate Use Claude Code for daily development tasks
Intermediate Know and use /compact, /cost, /model, /review
Intermediate Have a CLAUDE.md file in your projects
Intermediate Use Cowork for file automation and content creation
Intermediate Understand git basics: clone, branch, commit, push
Intermediate Switch between models based on task complexity

Level 3: Advanced

Advanced Use Plan Mode before every significant task
Advanced Create custom slash commands for repeated workflows
Advanced Connect MCP servers for external tools
Advanced Manage token costs with /compact, model switching, and budget caps
Advanced Use /rewind to safely experiment and roll back
Advanced Create custom subagents for specialized tasks

Level 4: Expert

Expert Orchestrate Agent Teams for parallel development
Expert Build skills with YAML frontmatter and tool restrictions
Expert Use hooks to automate pre/post-action workflows
Expert Configure permission modes for different contexts
Expert Optimize CLAUDE.md for minimal token usage with maximum context
Expert Use Claude Code SDK to build custom AI-powered tools
⚡ Where Is Damara?

Let's be real — you're solidly Intermediate heading into Advanced. You use Chat, Cowork, AND Code. You manage a GitHub Pages site. You understand tokens and costs. You're building branded content and pushing to production. The next steps: Plan Mode as a habit, custom commands, and MCP server connections. You're closer to Expert than you think.

18

Token Usage Across Modes

Not all modes are created equal when it comes to token consumption. Here's what each mode actually costs.

ModeToken UsageWhy
Chat (simple Q&A)Low (200-1,000 tokens)Short exchanges, no tool use
Chat (with files)Medium (1,000-10,000)File content counted as input tokens
Cowork (task)High (5,000-50,000)Multi-step planning and execution
Code (simple edit)Medium (2,000-8,000)Reads files, makes changes
Code (feature build)High (10,000-100,000)Reads codebase, writes multiple files
Code (agent teams)Very High (50,000-500,000+)Multiple parallel context windows

Where tokens go in Claude Code

Token breakdown for a typical Claude Code session: ████████████ File reading (35%) — reading your codebase ██████████ Conversation history (28%) — re-sent every message ███████ Output/code writing (20%) ████ System prompt + CLAUDE.md (12%) ██ Tool definitions (5%)

How to minimize token costs

  1. Use /compact regularly — compresses conversation history (the 28%)
  2. Keep CLAUDE.md lean — cuts the 12% on every message
  3. Switch to Haiku for simple tasks — same tokens, much cheaper per token
  4. Start new sessions for new tasks — don't carry unnecessary history
  5. Use Plan Mode first — avoids expensive false starts
  6. Be specific in prompts — vague prompts cause Claude to read more files
  7. Check /cost periodically — awareness prevents surprises
✅ The Max Advantage

On Max 20x, you get ~900 messages per 5-hour window. That's roughly 180 messages per hour. Even with heavy Claude Code usage, most developers never hit that limit. And if you do, Extra Usage kicks in at API rates — you're never cut off.

🏆 10 Golden Rules of Claude Mastery

1
Use the right mode for the job. Chat for thinking, Cowork for desktop automation, Code for development. Using Chat when you need Code is like bringing a knife to a gunfight.
2
Plan before you build. Enter Plan Mode with /plan before every significant task. Let Claude research and propose. Approve, then execute. This prevents expensive wrong turns.
3
Switch models by task complexity. Haiku for quick lookups. Sonnet for day-to-day coding. Opus for complex architecture. Using /model is the easiest way to balance quality and cost.
4
Run /compact every 15-20 messages. Long conversations bloat your context and cost more with every message. Compacting is free and saves you tokens and money.
5
Create a CLAUDE.md for every project. It's like onboarding a new team member — once. Include tech stack, coding standards, key commands, and important conventions. Keep it lean (tokens add up).
6
Build custom commands for repetitive workflows. If you do something more than twice, make it a /command. Deploy scripts, review checklists, test routines — automate the patterns.
7
Use /cost and /context as habit. Check your token usage and context fill level regularly. Awareness prevents budget surprises and context overflow.
8
Be specific in your prompts. "Fix the login bug in src/auth/login.ts" uses fewer tokens than "something is wrong with login." Specificity means Claude reads fewer files and makes fewer guesses.
9
Use /rewind as your safety net. Experimentation is cheap when you can roll back. Try bold approaches knowing you can always rewind to a known good state.
10
Claude is your teammate, not a magic box. Give context. Review output. Iterate. The best results come from treating Claude like a senior developer you're pairing with — not a genie you make wishes to.