Shiny Things

Company: Shiny Things

Self-reported basis: The entire analysis is based on the author’s own description of the project, submitted to the OpenAI 2026 hackathon on Devpost. No external verification or historical data are available.

What it appears to be: A values-grounded attention tool that uses AI to surface discrepancies between a user's stated values and their actual behavior, without making decisions for the user. The app is built with AI (GPT-5.6) and aims to help users reflect on what matters most, rather than simply capture tasks or manage time.

What changed: The project was submitted as part of a hackathon, suggesting it is in early development or prototype form. It does not appear to have launched publicly or gained traction beyond the submission context.

Single most important open question: Is there evidence that users will engage with the tool regularly enough to find value in its reflective, non-coaching approach — especially when it surfaces uncomfortable truths?

Confidence level: Low. The description is self-reported and unverified. No data on usage, adoption, revenue, or customer feedback are available.

The description states that Shiny Things is a values-grounded attention tool. It allows users to define real values with real definitions and what makes them hard to live. The app watches the gap between what users say matters and where their attention actually goes.

It uses an AI witness layer, powered by GPT-5.6, that surfaces these gaps without coaching or making decisions for the user. Every AI suggestion is sourced, and users can see why something was flagged.

The tool is built with:

• Codex
• GPT-5.6
• Next.js
• Tailwind CSS
• TanStack Query
• TypeScript
• Zod
• Zustand

Inference: The product appears to be a prototype or early-stage tool focused on reflective self-management, not task management or productivity optimization.

The author states that the product is built around the question: “Is this important to me—or is it only shining loudly right now?”

It positions itself as an alternative to traditional productivity tools that reward action over discernment. The core idea is to help users distinguish between what they say matters and where their attention actually goes.

The author also mentions a deeper inspiration from patterns in medicine, leadership, and relationships — the gap between institutional performance and actual delivery.

Inference: The positioning is rooted in philosophical and ethical concerns around AI use in personal tools. It claims to avoid manipulation or dependency by design.

The description does not name specific customer segments or personas. However, it implies a user who:

• Values self-reflection
• Is interested in aligning behavior with stated values
• May be frustrated with traditional productivity tools that don’t support discernment
• Is open to AI that surfaces uncomfortable truths without coaching

Inference: The ICP likely includes individuals seeking personal development or mindfulness, possibly in professional or creative contexts. It is not clear if this is a broad audience or niche.

No business model or pricing information is provided in the description.

Not evidenced: There is no mention of monetization, subscriptions, freemium models, or any revenue streams.

The app is built with:

• Codex
• GPT-5.6
• Next.js
• Tailwind CSS
• TanStack Query
• TypeScript
• Zod
• Zustand

It uses strict TypeScript, typed domain models, and explicit provenance on AI suggestions.

The AI witness layer is designed to:

• Not override user decisions
• Show why a suggestion was made
• Allow users to control what the AI can see (analysis permission model)

Inference: The technical stack suggests a modern web application with strong type safety and AI integration. The design philosophy emphasizes transparency and user agency.

The project was submitted to the OpenAI 2026 hackathon, indicating it is in early development or prototype form.

There is no evidence of:

• Public launch
• Users or customers
• Revenue or monetization
• Product adoption or retention metrics

Not evidenced: No traction data, usage statistics, or customer feedback are available.

The description does not mention competitors. However, the core idea — using AI to reflect on attention and values — aligns with:

• Mindfulness apps
• Personal development tools
• Attention management platforms
• AI-powered journaling or reflection tools

Inference: The product may compete with tools that focus on personal growth or mindfulness, but it is not clear how it differentiates from existing offerings in the market.

1. Lack of user engagement model: The tool surfaces uncomfortable truths without coaching or guidance — this may reduce user retention unless users are highly self-motivated.
2. No monetization strategy: No evidence of a business model, which raises questions about long-term viability.
3. High philosophical constraints: The product’s design is deeply rooted in moral and philosophical principles, which could limit scalability or flexibility.
4. AI tone challenges: The description notes that writing an AI that surfaces uncomfortable truths without sounding like a coach or judge is hard — this may affect user experience.

1. What specific values do users define in the tool, and how are these defined?
2. How does the app handle edge cases where a user’s stated value conflicts with their behavior in complex ways?
3. Are there any plans to test or validate the product with real users beyond the hackathon context?
4. What is the intended path from prototype to product — is this meant to be a long-term tool or a short-term experiment?
5. How does the app ensure that AI suggestions are not misinterpreted or misunderstood by users?