- What is Claude
- Claude models
- How chats work
- Context window
- Best use cases
- Pricing & plans
- Interface overview
- Settings to know
- Keyboard shortcuts
- Mobile & apps
From beginner to power user.
A structured roadmap through the Claude surface area — from your first prompt to building multi-agent systems. Ten stages, one hundred concrete skills, and the frameworks that connect them. No fluff, no shortcuts — just the path.
The 10-Stage Map
- Role prompting
- Context stacking
- Chain-of-thought
- Add constraints
- Prompt chaining
- Critique outputs
- Rewrite workflows
- System prompts
- Reusable templates
- Iteration loops
- What are Artifacts
- When they trigger
- Artifact prompting
- Simple projects
- Responsive design
- Live editing
- Debugging
- Frontend basics
- React prompting
- Mini products
- What are Skills
- Skill structure
- Writing Skills
- Research Skills
- Strategy Skills
- Skill libraries
- Consistency
- Reduce hallucination
- Workflow specialisation
- Stop starting from scratch
- What is Coworker mode
- Collaborate
- Reviewer
- Strategist
- Researcher
- Editor
- Teacher
- Planner
- Multi-role sessions
- Operational support
- What are Connectors
- Google Drive
- GitHub
- Notion
- Slack
- File analysis
- Spreadsheet analysis
- Centralised workflows
- Permissions setup
- Access your data
- What is MCP
- Why MCP matters
- Agent loops
- Automation chaining
- API basics
- Zapier
- Make
- Memory systems
- Guardrails
- Autonomous runs
- Claude Code overview
- Explain code simply
- Debug errors & fixes
- Refactor code
- Full-stack projects
- Write documentation
- Unit tests & analysis
- Best practices
- Dev workflow tips
- Ship better code
- Marketing & content
- Sales enablement
- Customer support
- Research & analysis
- Product strategy
- Operations & HR
- Finance & reporting
- Legal & compliance
- Education & training
- Personal productivity
- Advanced prompting
- Custom workflows
- Multi-agent systems
- Advanced integrations
- Scalable systems
- Performance tuning
- Deep research ops
- Business automation
- Build your systems
- Operate the stack
Frameworks & Quick Reference
- Role
- Who Claude should be
- Context
- Background, audience, situation
- Task
- What Claude should do
- Constraints
- Rules, tone, length, style
- Format
- How the output should look
- WSWebsite Landing pages
- DBDashboards Analytics
- APMini apps Internal tools
- CLCalculators Logic
- FMForms Surveys
- DCDocuments Long-form
- Code Frontend
- DGDiagrams Flowcharts
- Role
- Definition of the expert
- Rules
- Hard guardrails
- Workflow
- Step-by-step process
- Examples
- Concrete clarifications
- Format
- Output contract
- BRBrainstormer
- STStrategist
- RVReviewer
- RSResearcher
- EDEditor
- TCTeacher
- PLPlanner
- PSProblem solver
- GDGoogle Drive Files & Docs
- GHGitHub Repos
- NONotion Workspace
- SLSlack Conversations
- WBWeb search Live info
- MCPMCP servers External tools
- APIAPIs Integrations
Operational Workflows
Topic / niche
Brief and audience
Claude gathers
Sources, data, quotes
Structured frame
Sections & hierarchy
First write
Full content draft
Critique & rewrite
Refine and tighten
Multi-channel
Reels, posts, blog
Schedule & ship
Distribute
File / form / event
Connected system
Analyse / generate
Reasoning + tool use
Result / summary
Structured response
Send to tools
Slack, Notion, CRM
Tips, Commands & Pitfalls
- Be specific
- Add context
- Use examples
- Break big tasks
- Iterate outputs
- Use formatting
- Save prompts
- Build workflows
/explainAnything simply/summarizeLong content/rewriteWith improvements/improveSharper version/shortenMake concise/expandAdd detail/compareTwo things/analyseDeep analysis/translateAny language
- Act as a [ROLE] expert.
- I’m working on [TOPIC].
- My goal is [GOAL].
- My audience is [AUDIENCE].
- Rules: [CONSTRAINTS].
- Output in [FORMAT].
- Think step-by-step before output.
- Vague prompts
- No context provided
- Asking too much at once
- Not iterating
- Ignoring output formatting
- Starting new chats too often
- Not using examples
- Not saving prompts
- Underusing Artifacts & Skills
- Trusting outputs blindly
Level-Up Path
From learner to operator
Read the full article or download as PDF
The full article and the PDF are member content. Magic-link login, no credit card, no risk — both available immediately.