VibeMap Logo

Creating Your First Project

On this page

Creating Your First Project

This guide walks you through creating a new project in VibeMap, from filling out the project form to reviewing your AI-generated project summary.

Step 1: Open the New Project Form

From the dashboard, click the New Project button. This opens the project creation form where you'll describe your product idea.

New project form

Step 2: Fill in Project Details

The new project form has two fields:

Project Name

Give your project a short, descriptive name. This is how it will appear in your dashboard and throughout the app.

  • Keep it concise (2-5 words)
  • Use the product name if you have one, or a descriptive label
  • Examples: "TaskFlow", "Recipe Sharing App", "Freelancer CRM"

Project Description (Prompt)

This is the most important field. Write a natural-language description of the product you want to build. The AI uses this description as the foundation for everything it generates.

A good description typically covers:

  • What the product does -- Its core purpose and value proposition
  • Who it's for -- The target users or audience
  • Key features -- The main things users can do
  • Technical preferences -- Stack, platform, or integration requirements (optional)

You don't need to be exhaustive. A few well-written paragraphs are better than a vague essay. See Writing Effective Prompts for detailed guidance.

Step 3: Choose a Project Mode

VibeMap offers different project modes that tailor the generation pipeline to your needs:

  • Full Spec -- Generates the complete set of deliverables: personas, features, user stories, acceptance criteria, pages, schema, and file structure. Best for new projects that need comprehensive planning.
  • Quick Start -- Generates a focused subset of deliverables to get you started fast. Useful for prototyping or validating an idea before committing to a full spec.

Select the mode that matches your current goal. You can always generate additional deliverables later.

Step 4: Submit and Wait for Generation

Click Create Project to submit. VibeMap's AI pipeline begins processing your description:

  1. The AI analyzes your prompt and generates a project summary -- a structured overview that captures your product's goals, target audience, core value proposition, and scope.
  2. This typically takes 15-30 seconds depending on the complexity of your description.
  3. You'll see a progress indicator while the AI works.

Once complete, you'll be taken to the project summary view.

Step 5: Review the Project Summary

The project summary is the AI's interpretation of your product idea, organized into clear sections:

Project summary

  • Overview -- A concise description of the product and its purpose
  • Target Audience -- Who the product serves and their primary needs
  • Core Value Proposition -- What makes this product useful or different
  • Key Goals -- The main objectives the product should achieve
  • Scope -- Boundaries of the initial version

Review carefully

The summary sets the direction for all downstream generations. If the AI misunderstood something:

  • Edit the summary directly to correct any misinterpretations
  • Regenerate the summary if the overall direction is off
  • Refine your prompt and create a new project if needed

Getting the summary right is worth the time -- it cascades into better personas, features, and everything else.

Step 6: Start Generating Deliverables

With your summary confirmed, you're ready to generate the full set of project artifacts. Navigate through the project tabs to generate each deliverable in sequence:

  1. Business case — AI-generated market validation (optional, Pro+)
  2. Personas — Target user profiles
  3. Features — Core product functionality
  4. User stories — Requirements from the user's perspective
  5. Acceptance criteria — Testable definitions of done
  6. Pages & sections — UI structure and routes
  7. Database schema — Data model and relationships
  8. File structure — Directory layout and key files

Each step builds on the previous ones, so generating them in order produces the most coherent results.

Step 7: Hand off to your AI coding agent

Once your spec is in shape — features approved, acceptance criteria reviewed — the next step is to connect your IDE's AI coding agent (Cursor, Claude Code, Windsurf, VS Code Copilot) to VibeMap via the MCP server. Your agent picks up work one acceptance criterion at a time, implements it in your codebase, and submits for review.

Full walkthrough → Handoff to your AI agent.

Tips for Your First Project

  • Start simple — Try a straightforward product idea for your first project to learn how the pipeline works
  • Don't skip the summary review — Fix any issues before generating downstream artifacts
  • Experiment freely — You can regenerate any deliverable as many times as you need
  • Use the agent — The conversational AI agent can help you refine specific parts of your spec

Next Steps