Ignis Algos

Snap a picture of an area you want to clean up — TidyQuest turns it into a guided quest with a clay companion, and every finished quest earns a gift for its nook. Snap a picture of an area you want to clean up. Meta AI analyzes the photo, and TidyQuest turns the space into a guided cleanup quest with a clay companion at your side. Finish quests to earn gifts for your companion's nook, a golden star medal, and the cape. No fail states. No streaks. No shame. TidyQuest — tidying for the days when the mess wins before you even start. THE PROBLEM An overwhelming space — a room, a desk, a car interior — is hardest at the moment of starting. Most tidying tools hand you a generic checklist, a streak to protect, or a before-photo and a guilt trip. For people whose energy is limited or unpredictable, that design fails exactly when it is needed most: the streak breaks, the plan does not fit the day, and the tool becomes one more thing to feel behind on. WHAT TIDYQUEST DOES Snap a picture of an area you want to clean up. That photo is the whole setup. Meta AI analyzes the photo, and TidyQuest turns the space into a guided cleanup quest: the areas of the space become stops on a quest map, the biggest area becomes the Boss, and a clay companion — one of the clay creatures, Robot, Blob, Star, and Sprout — walks the quest with you. You work the quest at your own pace, area by area. There is no clock on you, and no way to fail a quest. Finishing quests earns gifts, and the gifts are real in the world of the app: your companion's nook fills up, gift by gift — a book, a mug of cocoa, a comfy chair — on the way to a golden star medal, and finally the cape, the celebration a full clear deserves. The nook is where your companion lives between quests, surrounded by everything you earned together, and it stays lively: your companion reads the book, sips the cocoa, and admires the medal it earned. THE DIGNITY PRINCIPLE (Impact) There are no fail states, no streaks to lose, nothing counting down. Rest is a legitimate move. The plan bends to the person's day, never the other way around. TidyQuest is built for the people tidying apps usually abandon: anyone overwhelmed, energy-limited, or starting from a space that feels impossible. The real-world potential is a tidying companion people keep using precisely because it never punishes them. HONEST BY DESIGN TidyQuest never inflates a result. A finished quest is celebrated as exactly that — the quest you set and completed, presented in quest mode. The app does not dress up a counted before/after reduction it cannot stand behind. What it promises is a guided, finishable path through a real space, and it keeps that promise. USE OF META AI TOOLING Meta AI is the brain of the quest, reached through the Meta Model API — detailed in the tooling section below. TECHNICAL DEPTH & EXECUTION A Flutter Android app backed by a hosted proxy: the app sends the photo to tidyquest.ignisalgos.com, the proxy calls the Meta Model API, and the structured response becomes the quest map, the steps, and the companion's guidance. The judge endpoint is built into the shipped app — no account, no API key, and no special network is needed to run it. The companions are rigged 3D clay characters rendered as animation loops, and the nook, the gift system, the medal, and the cape are earned state that persists between sessions. Scope is deliberately bounded: one space per quest. No whole-house mode in v1. Meta AI does the thinking at the top of the loop; everything else in TidyQuest is built around its answer. The app photographs a space and sends the photo to our hosted proxy. The proxy calls the Meta Model API (api.meta.ai/v1): the model analyzes the space and returns a structured read of it — the distinct areas in the photo and what is in them. TidyQuest turns that structure into the product: areas become stops on the quest map, the largest area becomes the Boss, and the step-by-step guidance and the companion's wording are generated from the analysis rather than from a generic room template. Behind that single API surface, the pipeline stages use Meta's models for the jobs each is best at — segmentation of the photo into the areas and items of the space, planning that turns the analysis into a sequenced quest, and a generated vision of the tidied space the user can open mid-quest. The live path was verified end to end through the same public endpoint the shipped app uses. The API key lives only on the server. The app ships with the endpoint built in, so there is no key in the app or the repository, and judges need no key of their own. Judges — running TidyQuest takes an Android phone and nothing else: 1. Download the APK: https://truegradefcg.com/tidyquest/TidyQuest1.4.7.apk (v1.4.7, the final build — see the links below; if a newer versioned link appears there, use it.) 2. Install it (Android may ask you to allow installs from your browser) and open TidyQuest. No account, no sign-in, no API key — the app is preconfigured to reach our hosted endpoint at tidyquest.ignisalgos.com. 3. Pick your companion: Robot, Blob, Star, and Sprout are ready. (Jane, the fifth clay character, fronts the guided tour on the home screen.) 4. Snap a picture of an area you want to clean up — a desk, a shelf, a car seat, the corner of a room. Meta AI analyzes the photo and builds your quest map. 5. Work the quest area by area, at whatever pace feels right. Finish the areas to finish the quest. 6. Collect what you earn: gifts appear in your companion's nook, and a full clear earns the golden star medal and the cape finale, celebration included. The endpoint stays live through the judging window (Nov 4-8). A demo video accompanies this submission — see the links below. GitHub repository (public): https://github.com/parabellumlux/tidyquest Downloadable build (Android APK): https://truegradefcg.com/tidyquest/TidyQuest1.4.7.apk Build version: v1.4.7 (final build — built and parent-verified Nov 2: 86/86 tests, hosted download byte-identical, 33,044,554 B, MD5 af5d7a396d9aa6ada7528e18833bc010) Live judge endpoint (built into the app): https://tidyquest.ignisalgos.com Demo video: [VIDEO LINK: PENDING — paste the YouTube URL here before filing] TidyQuest has no accounts and keeps no profiles. When you snap a photo, the app sends it to our hosted proxy at tidyquest.ignisalgos.com, whose only job is to pass it to the Meta Model API for analysis and return the quest — that proxy is the app's only network destination, and its only upstream is Meta's API. No API key is embedded in the app or the repository; the key lives on the server. Demo materials use only spaces their owners cleared for the demo, and no family data appears in the repository, the video, or this submission. Team / Startup: Ignis Algos Track: Startup Team Representative (submits on behalf of the team): Damon Pettigrew — co-founder and an active member of Ignis Algos, duly authorized to submit this project and act on behalf of the team, and the team's designated individual to receive notices. Team member: Diana Pettigrew Each member of the team has read the Hackathon Official Rules and agrees to abide by them. TidyQuest was designed, built, and tested by this two-person team, working with their AI agents (Remy and Baymax) throughout the event. TidyQuest is the original work of the Ignis Algos team and is solely owned by the team. Open-source components are used under their applicable licenses, and the submission builds on them as permitted.