Mind Mirror

Inspiration

Mind Mirror was inspired by rubber duck debugging, the practice of explaining your problem out loud to a rubber duck until the solution becomes clear.

We realized writer’s block is rarely about lacking ideas. It is about unclear thinking. When you speak your thoughts out loud, your brain reorganizes them. A rubber duck helps you process, but it does not ask follow up questions.

Mind Mirror became an AI powered reflective companion that does not give you answers but helps you uncover them yourself.

What it does

Mind Mirror is an AI writing companion designed to break creative blocks through structured and adaptive questioning.

Users choose from three personas, Sage who is reflective, Manager who is direct, and Detective who is curious. They fill out a structured intake form about their project and begin a conversation that progresses from surface ideas to deeper insight.

The system supports voice input for natural thinking, generates a session summary, and visualizes themes through an interactive mind map.

Instead of generating content, Mind Mirror generates clarity.

How we built it

We built Mind Mirror using Vite for fast iteration, React 18 and TypeScript for a scalable frontend, Tailwind CSS and shadcn ui for accessible components, and ReactFlow for dynamic mind map visualization.

The conversational backend runs through OpenRouter API to free models.

We implemented persona based system prompts, progressive questioning logic that moves from general to specific reflection, and structured insight extraction to power summaries and mind maps.

The architecture separates conversation logic from visualization logic so the system remains extensible.

Challenges we ran into

One major challenge was avoiding generic AI responses. We had to carefully refine prompts to ensure the system asked layered and meaningful questions rather than surface level ones.

We also struggled with repetitive questioning and added user controls to redirect or pause the flow.

Another challenge was translating open ended conversation into structured mind maps. Extracting themes and relationships from free form text required thoughtful summarization logic.

Finally, we had to balance guidance with autonomy so the AI supported thinking without taking over the process.

Accomplishments that we're proud of

We built a system that strengthens creativity rather than replacing it.

We designed three distinct personas with noticeably different cognitive styles.

We created a calm and intentional user experience that makes deep thinking feel accessible.

We successfully integrated conversation, summary, and visualization into one seamless workflow.

Most importantly, we built something that feels genuinely helpful rather than gimmicky.

What we learned

Writer’s block is often a clarity problem, not a creativity problem.

The quality of questions matters more than the quality of answers.

Structured reflection accelerates ideation.

Strong AI products are built on augmentation rather than automation.

Simple interfaces can enable deep thinking.

What's next for Mind Mirror

We plan to build long term memory across sessions and project based insight tracking.

We want to generate exportable writing briefs from conversations and introduce collaboration mode for shared reflection.

We aim to personalize questioning styles based on user thinking patterns and improve semantic clustering for richer mind maps.

Our long term vision is to make Mind Mirror the default thinking companion for anyone navigating the space between idea and execution.

Built With

• javascript
• lovable
• openrouter
• react
• typescript