Daily stoic reflection in under two minutes - grounded in your actual day.
Aurelian helps you align your day, stoically, in under two minutes. The app combines a personal 12-week focus goal, the day's most important tasks, your current energy level and vetted stoic quotes into a short, concrete daily reflection. Deliberately not another quote-of-the-day app, but a stoic perspective on the user's actual day - calm and grown-up in tone, without motivational platitudes.
Stoicism apps serve generic quotes with no bearing on everyday life, journaling apps demand too much input and feel like work, and AI reflection tools invent quotes - with philosophical sources that destroys credibility instantly. Ambitious users want a short, practical moment of clarity before the day: no to-do list, no coach tone, no made-up wisdom.
Everything end-to-end - product definition and PRD, a modular local-first architecture, the SwiftUI frontend, domain and application logic as a Swift package, curating the vetted quote database, the hardened Node backend endpoint for the AI generation, plus deployment, App Store preparation and test strategy.
The core is a very short daily loop. In the morning the user enters one to three important things, optionally adds what is on their mind, and picks their energy level; from that Aurelian generates a stoic perspective and a concrete rule for the day. In the evening, two or three short questions follow. The output is strictly structured, and the AI only ever selects from vetted, locally supplied quote candidates.
Local-first and modular, with clear boundaries between product rules and technical details (UI, persistence, notifications, AI provider). A platform-independent Swift package is cut into five library targets (Domain, Application, Content, Infrastructure, Design); the SwiftUI app composes them. Clear boundaries are preferred over heavy abstractions.
The central trust question: if the AI invents quotes, authors or sources, that instantly destroys credibility for philosophical texts.
A vetted local database plus a deterministic QuoteSelectionPolicy hands the AI only permitted candidates. The AI picks exactly one ID but never invents - provenance (author, work, reference) stays guaranteed real.
Standard prompts produce interchangeable motivational lines with no bearing on the user's actual day.
A very strict prompt with real daily data in context and short, concrete outputs instead of motivational speeches - a stoic perspective on this exact day.
Direct in-app calls to the LLM provider expose the key and make costs uncontrollable.
Moving generation to a stateless server endpoint with a server-side key, fixed-window rate limiting, limited retries and exactly one output per user per day.
SwiftUI and SwiftData are hard to test in isolation - product rules risk being chained to the UI framework.
The modular, port-based architecture keeps domain and application logic independent of SwiftUI/SwiftData and thus testable across more than 26 test files in five test targets.
Domain package, content validation, daily-entry rules, design tokens, first application use cases (onboarding, morning check-in), SwiftData repositories and a SwiftUI app shell with onboarding, check-in, journal and focus wired up.
Deeper persistence, extended onboarding and journal detail; then the deterministic quote-selection policy, the catalog expansion, the decision for Groq, a remote reflection adapter, the local notification scheduler and first tests.
Stateless Groq endpoint implemented, hardened, deployed to production and validated by iOS E2E; plus App Store signing, shell assets and the public marketing, privacy and support pages.
App Store compliance answers, a store-metadata draft, polishing the accessibility copy and buffering the text inputs. Status since: Client MVP Ready / App Store Release Prep.