Claude Code for Non-Developers Training Course

Phase 1 — Introduction to Claude Code — 55 minutes

• Defining Claude and distinguishing Claude Code from standard chat interfaces
• Overview of the Claude ecosystem: claude.ai, Claude Desktop, and Claude Code (CLI), including their respective roles
• Interface navigation: exploring the Claude app, initiating coding sessions, and mastering the workspace
• Cognitive workflow: understanding the describe → plan → act → review cycle
• Permissions management: understanding why Claude requests approval before generating files or executing code
• Initial project: instructing Claude to generate a styled webpage from a single-sentence description
• Iterative refinement: implementing adjustments such as resizing headers, altering color schemes, or adding navigation elements
• Guided practice: Participants will open the Claude app, initiate a Claude Code session, and create a personalized “About Me” page by describing their requirements in plain English, then refine the output using follow-up commands.

Objective: Ensure all participants are familiar with the interface and have successfully completed their first interaction.

Break — 10 minutes

Phase 2 — Constructing Practical Applications with Natural Language — 70 minutes

This segment forms the core of the morning, focusing on completing four tasks of increasing complexity using only natural language prompts.

• Task 1 — Interactive Dashboard: Request Claude Code to create a styled dashboard displaying sample data with charts, statistics, and a clean layout. Practice providing design directives such as “apply a dark theme,” “add a sidebar,” or “ensure responsiveness.”
• Task 2 — Data Analysis: Upload a sample CSV file and instruct Claude to summarize the data, identify trends, extract minimum and maximum values, and generate a visual chart. This demonstrates Claude’s ability to write and execute code autonomously.
• Task 3 — Document Generation: Instruct Claude to read a data file and produce a formatted report, such as a sales summary, project status update, or meeting recap, illustrating the transformation of raw data into polished deliverables.
• Task 4 — Automation Utility: Ask Claude to build a simple interactive tool, such as a unit converter, quiz application, or budget calculator, highlighting its capability to create dynamic tools beyond static pages.

Following each task, the instructor will dissect the backend processes, including file creation, code generation, and output interpretation. Participants will log their most effective prompts in a shared Prompt Playbook.

Break — 10 minutes

Phase 3 — Optimizing Claude Code Usage — 50 minutes

• Prompting strategies: differentiating between specific and vague instructions
• Live demonstration: side-by-side comparison of ineffective and effective prompts applied to the same task
• Refinement techniques: requesting explanations for Claude’s choices, reverting changes, or exploring alternative approaches
• File integration: processing uploaded documents for summarization or converting spreadsheets into visual charts
• Multi-step workflows: chaining requests to generate complex outputs, such as analyzing data and subsequently building a dashboard
• Resource management: understanding tokens, context windows, and subscription tier implications
• Selection criteria: determining when to use Claude Code versus standard Claude chat
• Guided practice: Participants will extend a Phase 2 project by adding two new features using a multi-step prompt chain, then compare their prompts to identify factors contributing to improved results.

Objective: Progress from basic functionality to consistently generating high-quality results.

Break — 10 minutes

Phase 4 — Collaborative Workflow Construction: Live Build Session — 60 minutes

This phase shifts the dynamic from individual practice to collective problem-solving. The instructor facilitates the session while participants contribute real-world challenges from their professional roles, suggest prompt ideas, and debate trade-offs. The aim is to develop prompt judgment by observing skilled navigation of uncertainty in real time.

The session is structured around three workflow archetypes:

• Transform — Converting input X into output Y (e.g., meeting notes to action items, raw data to summary emails, or customer feedback to themed reports)
• Draft — Generating initial versions of documents typically created from scratch, such as proposals, emails, job descriptions, or social media posts
• Analyze — Interrogating complex documents or datasets that lack time for detailed review, such as lengthy reports, survey spreadsheets, or contracts

Setup and framing (10 min): The instructor introduces the archetypes and explains the session mechanics. Participants submit real workflow challenges from their roles via a shared document or chat.

Live build #1 — Transform workflow (20 min): The instructor selects a submitted problem and builds a solution live, incorporating prompt suggestions, pushbacks, and refinements from the room. Every decision is narrated. The session concludes with a working prompt template retained by the participant who submitted the problem.

Live build #2 — Draft or Analyze workflow (20 min): Similar format, applying a different archetype to a different participant’s problem.

Reflection and feedback (10 min): Participants reflect on surprising prompting tactics, areas for improvement, and key takeaways. A brief group share is facilitated, and the instructor connects observations to the broader Prompt Playbook.

Phase 5 — Integrating Claude with External Tools via MCP — 50 minutes

• MCP (Model Context Protocol) explained: the universal integration framework for AI tools
• The value of MCP: evolving Claude from a chat assistant into a connected workflow hub
• Connectors Directory: browsing and adding integrations directly within the Claude app
• Desktop Extensions: one-click installations for Claude Desktop requiring no manual configuration

Live demonstration: The instructor connects Claude to two services via the Connectors UI and demonstrates cross-tool workflows:

1. “Retrieve tomorrow’s meetings from Google Calendar and draft a preparation email for each”
2. “Read the latest updates from our project board and compile a status summary”
3. “Extract data from a connected service and generate a local report”

Guided practice: Participants connect Claude to at least one service, with options provided for varying comfort levels:

• Option A: Connect a pre-built connector from the directory (e.g., Gmail, Google Drive, or a demo service) via click-to-authenticate
• Option B: Add a custom connector by pasting an MCP server URL (a test URL is provided by the instructor)
• Option C: Install a Desktop Extension from the marketplace (for Claude Desktop users)

Participants then assign Claude a task utilizing the connected service, such as “Summarize my recent emails regarding project updates into a document.”

Key concepts covered:

• Connector mechanics: OAuth authentication, permissions, and access scope
• Access management: enabling, disabling, and controlling which connectors are active per conversation
• Security best practices: connecting only to trusted services and reviewing tool permissions
• MCP ecosystem: sources for new connectors, extensions, and community-built servers

Objective: Recognize Claude as a connective layer across existing services, not just a coding utility.

Break — 10 minutes

Phase 6 — Capstone Project and Future Steps — 65 minutes

Capstone mini-project (45 min): Each participant selects one scenario to build with Claude:

1. A polished landing page or portfolio site for a team, project, or personal brand
2. A data analysis pipeline: uploading a file, having Claude analyze it, and generating a visual report
3. An interactive tool addressing a real workflow problem (e.g., calculator, tracker, converter, or quiz)
4. A connected workflow: extracting data from a service, transforming it, and producing a deliverable (e.g., “Create a visual schedule from next week’s calendar”)

The instructor circulates to assist with prompt refinement and highlights standout examples to the group.

Showcase and conclusion (20 min):

• 6–8 participants present their projects (2–3 minutes each)
• Next steps: Claude Code CLI for terminal users, VS Code extension for developers, and Cowork for knowledge workers
• MCP ecosystem exploration: finding and evaluating new connectors, extensions, and community servers
• Subscription plans: Free vs. Pro vs. Max — features and suitable use cases
• Best practices review: recap of effective Prompt Playbook patterns from the session
• Recommended resources: official documentation, community channels, and Anthropic’s prompt engineering guide
• Participants receive a reference card detailing key prompting patterns, connector setup instructions, and a curated list of useful MCP integrations