Roadmap Ideas

Roadmap / Ideas

Everything on this page is tentative. LULL has no issue tracker or formal roadmap document in the repo. What follows is only what the repo's own status notes (README "Status" section, CLAUDE.md, the concept brief, and the fear-mechanics table) genuinely imply is next — not a commitment, not a schedule, and not anything beyond what's already written down somewhere in the project.

What v0.1 says is done

Per the README's own "Status" section:

LULLKit (consent, the EyeSession mechanic, and the Atmosphere voice) is tested and green — 16 tests, green from commit one; the SwiftUI + AVFoundation app in app/ implements THE EYE, consent-gated, speaking in the three registers above. [...] Not scary yet — but it watches, it has a voice, and it never breaks the rule.

Two things worth noting precisely: the project's own words are "not scary yet" — v0.1 is described as a proven mechanic and a working consent foundation, not a finished horror experience. And the vertical slice's purpose is explicitly a go/no-go test, per the concept brief: "does it make one person, alone at night, put the phone face-down? If yes, everything else is worth building. If no, we learned it cheap." No result of that test is recorded in the repo.

Implied next mechanics

The README's fear-mechanics table lists five more mechanics beyond THE EYE that are named and designed but have no code yet (see Concept & Design for the full table):

• THE ROOM — microphone, listening to silence and what breaks it
• THE REACH — a local notification at 3am, app closed
• THE HOUR — behavior that varies with how late/still it is
• THE PULSE — a haptic "heartbeat" that isn't quite the player's own
• BEHIND YOU — spatial audio placing something over the player's shoulder

Notably, LULLKit.Sensor already has cases for microphone, notifications, haptics, and motion — with honest rationale strings written for each — even though only .camera is wired into any actual session logic today. That reads as the allow-list having been designed ahead of the mechanics that will use it, which fits the "horror by permission" rule: the sensor has to be nameable and consent-gated before any mechanic can use it.

The haunt server

The README describes a server/ component, marked (later):

the Vapor haunt server: content that shifts between sessions, seems to respond, and is quietly shared between players.

There is no server/ directory in the repo yet. The concept brief frames this as reusing the Vapor/session/realtime stack from the MateMate chess project, calling it "already a solved problem" — but that's a stated intention to reuse prior work, not evidence of any LULL-specific server code existing.

A second mechanic, in ideation

THE MIRROR — a reflection that stops matching the player — exists as a design-notes document on an unmerged branch (ideation/mirror-and-still-here), explicitly marked "not a commitment." It's the most developed of the not-yet-built mechanics in that it has a written escalation ladder and proposed tech approach, but it has zero code and hasn't been merged to main. See THE MIRROR (ideation) for the full breakdown.

The SpriteKit question

Whether the Atmosphere experiment (PR #1, draft) gets adopted, replaced with a lighter SwiftUI/Core Animation approach, or left as a prototype is explicitly undecided — the PR frames it as evidence to look at, not a decision already made.

What's genuinely open, per the project's own framing

• Whether THE EYE actually scares anyone (the go/no-go test the whole slice was built to run) — no result recorded in the repo.
• Which, if any, of the other five mechanics get built next.
• Whether the SpriteKit prototype becomes the app's rendering approach for ambience, or purely reactive future mechanics use it while ambience stays in SwiftUI.
• The haunt server — design intent only, no code.
• Whether THE MIRROR gets built at all, and if so whether as part of THE EYE's slice or its own — the ideation doc leans toward "its own," but calls that undecided.