MinFo

Hook Page
Clarification: A gentle AI agent helps you talk through vague, scary ideas until they become concrete.
Atomization: It deconstructs your project into "atomic tasks"—physical actions that take less than 10 minutes—so you can't fail.
Sanctuary: You can arrange these artifacts in a virtual room and use AI to "Manifest" a high-quality visual representation of your progress
Gamified Reward

Inspiration MinFo was born from a shared frustration: most productivity tools feel like cold, demanding bosses. We wanted to build a "Quiet Mentor" that understands the anxiety of a blank page. Inspired by the Fogg Behavior Model, we realized that the biggest barrier to action is high cognitive load. We combined the "Zen" aesthetic of a foggy morning with the power of Large Language Models to create a space where productivity feels like a meditative journey, not a chore. What it does MinFo is a cognitive intervention system that helps users overcome procrastination through "Atomic Flow." Clarification: A gentle AI agent helps you talk through vague, scary ideas until they become concrete. Atomization: It deconstructs your project into "atomic tasks"—physical actions that take less than 10 minutes—so you can't fail. Gamified Reward: Completing tasks earns you "Stars," which you use to buy symbolic artifacts (e.g., a "Golden Compass" for navigation or a "Potted Wisdom" plant). Sanctuary: You can arrange these artifacts in a virtual room and use AI to "Manifest" a high-quality visual representation of your progress, creating a mental gallery of your mastery. How we built it We utilized a lean, high-performance stack to ensure the "Flow" is never interrupted: Core Logic: Built with React 19 and TypeScript for a fluid, type-safe user experience. AI Engine: We leveraged the @google/genai SDK. We used Gemini-3-Flash-Preview for its lightning-fast reasoning to deconstruct goals and Gemini-2.5-Flash-Image to manifest the visual sanctuaries. Styling: Tailwind CSS paired with custom Glassmorphism effects and the Patrick Hand typography to maintain a soft, non-threatening design language. Architecture: A stateless, ESM-based frontend that interacts directly with the Gemini API, featuring robust error handling and exponential backoff for API resilience. Challenges we ran into The primary challenge was balancing the AI's creativity with the user's focus. We had to strictly tune our System Instructions to ensure the AI remained a "Quiet Mentor"—listening more and talking less. Technically, integrating AI image generation required careful prompt engineering to translate abstract progress (like "completing a research paper") into evocative, architectural visual metaphors without breaking the app's serene aesthetic. Accomplishments that we're proud of We are incredibly proud of the "Manifest Vision" feature. It’s one thing to see a list of checkboxes; it’s another to see your hard work transformed into a beautiful, cinematic sanctuary. We also succeeded in creating a UI that feels "organic"—the mesh-gradient background and hand-drawn fonts successfully lower user heart rates, making it easier to start difficult tasks. What we learned We learned that "Lowering the Floor" is more important than "Raising the Ceiling." By using AI to automate the most painful part of any project—the planning and deconstruction—we found that users enter a "Flow State" significantly faster. We also learned that the tone of an AI (its persona) is just as important as its intelligence; users are much more likely to be honest about their struggles with a "Quiet Mentor" than a standard AI assistant. What's next for MinFo Stylized Manifestation: Moving from realistic photography to a 2D-illustration style that better matches our hand-drawn UI. Living Sanctuaries: Introducing the Gemini Live API to allow users to talk to their "Mentor" in real-time while they work. Collaborative Groves: Allowing teams to deconstruct giant projects into a shared "Grove" where everyone's progress contributes to a massive, collective architectural vision.