(conversation with Claude)
i need consultancy
i have this project:
## Project Overview
- Unit-testing and implementation project for LunyEngine (cross-engine C# for Unity, Godot, etc.).
- Folder structure:
- `Luny/`: Engine-agnostic types.
- `Luny-ContractTest/`: Engine contract verification tests of `*.*-Mock` types that simulate engine behaviour
- `Luny-Test/`: Tests for `Luny` and `LunyEngine` SDK features.
- `LunyLua/`: Lua-CSharp framework.
- `LunyLua-Test/`: Tests for `LunyLua`.
- `LunyScript/`: Scripting types.
- `LunyScript-Test/`: Tests for `LunyScript`.
- `*.Unity` / `*.Godot`: Engine-specific bindings using `*.*-Mock` types (this solution) or engine types (engine projects)
- `*.Unity-Mock` / `*.Godot-Mock`: Engine mocks/shims simulating native engine API and behaviour (this solution)
- `.junie/plan.md`: specific, step-by-step strategy for the current active task
- `.junie/tasks.md`: high-level roadmap / todo list
- `*.csproj` locations
- repository root (preferred): required for projects referenced in engine projects (eg Luny, Luny.Godot, Luny.Unity, ..)
- subfolders: for `*-Test/` projects
- LunyScript API surface is aimed at beginners and designers and should avoid technical noise and common beginner pitfalls.
- Luny.csproj is a developer SDK, it should still favor "friendly" semantics and designs but can be more technical.
It runs in Unity and Godot, and this solution which uses engine mocks/shims to simulate engine behaviour iterating across multiple engines at once albeit âflying blindâ and relying on AI to get the mock behaviour right. The architecture and basic functionality is implemented. Iâm practically ready to go an implement features necessary to make simple scripts and educational games.
Unfortunately, my first funding opportunity fell through. The next is due in a month, with a decision likely reached in four months. Itâs Epicâs Megagrant. Likely 4-10 times as many applicants with only a fraction being FOSS projects, but at least that fund is much better aligned with the game engine ecosystem.
Iâm completely self-funded, no backup plan. My savings would keep me afloat for 12 months reasonably, 18 if i stretch it to the limit. So I need to somehow make money in that time .. from a FOSS project. Or turn it commercial. Or half-half.
Since I have no other options I need to move ahead. And the focus should be on impressing Epicâs reviewers. They recommend to send in a Google Drive link so Iâll be able to update application material and links but I donât know when theyâll actually look at it.
I should probably try to generate some buzz around the project. It already reached 107 GitHub stars in 3 months, mostly thanks to an initial social media link posting boost but mainly due to my Unity Discussions (forum) signature and standing in that community (Top 10 poster). That rise of stars far outpaces any other FOSS repository of mine, the runner up and âprequelâ was a Lua integration in Unity which lead to this project (Lua integration is also a goal).
Naturally Iâm a bit bummed out. I have to stay motivated despite having limited time, burning through my money like Iâve done the past 3 years with practically nothing to show. Best thing now is to try and find a reasonable plan of attack to:
a) increase my chances of getting funded by Epic b) generally increase funding opportunities (Patreon, GitHub Sponsors, Kickstarter .. ?) c) allow me to pivot toward commercial distribution (Unity Asset Store, paid tutorials, services) d) exit strategy: get paid/employed by game or game tool developer, ideally utilizing my work in parallel
My actual motivation for this tool is to bridge engines, make them more interoperable on the code-level, converge towards a common standard, and make it easier for beginners and designers to utilize game engines and to make games faster, built more reliably, with higher code re-use, and without the necessary extra layers causing too much friction (performance, memory usage, debugging depth, etc).
what do you suggest we should talk about in more detail?
Given your situation, let me first understand the project better, then we can discuss strategic priorities.
Key areas to explore:
Given your 12-18 month runway and 4-month Epic decision:
Talk about first:
What specific help do you need from me?
Which of these areas feels most urgent to you right now?
please find my PrototypeFund application attached (it was dismissed, Iâll have to wait weeks for reasons)
Iâll answer your topics first:
I made a proof of concept prototype 3 months ago - it runs in Godot, Unity and Unreal (via UnrealSharp). Took 20 days to make. I have a GIF showcasing this on the website, and a video on Youtube. That code didnât implement a foundational cross-engine architecture however, I took some shortcuts as usual for a PoC. Best I have with current architecture is bouncing balls and a Xmas scenery with blinking lights and Kenney.nl assets.
As to âwowâ factors:
public sealed class PhysicsSphereSpawner : Script
{
public override void Build(ScriptContext context)
{
var instantiate = Prefab.Instantiate("Prefabs/PhysicsSphere");
var log = Debug.LogInfo("SPAWN SPHERE");
On.Ready(log, instantiate);
var createCounter = Counter("Sphere: Create").Every(50).Heartbeats().Do(log, instantiate);
Counter("SphereSpawner: Destroy").In(50).Heartbeats().Do(createCounter.Stop());
Timer("RELOAD").In(8).Seconds().Do(Scene.Reload());
}
}
public sealed class PhysicsCubeSpawner : Script
{
public override void Build(ScriptContext context)
{
var instantiate = Prefab.Instantiate("Prefabs/PhysicsCube");
var log = Debug.LogInfo("SPAWN CUBE");
Counter("Cube: Create").Every(70).Heartbeats().Do(log, instantiate);
Counter("CubeSpawner: Destroy").In(150).Heartbeats().Do(Object.Destroy());
}
}
public sealed class PhysicsSphere : Script
{
public override void Build(ScriptContext context) => Timer("Sphere: Destroy").In(7.3).Seconds().Do(Object.Destroy());
}
public sealed class PhysicsCube : Script
{
public override void Build(ScriptContext context) => Timer("Cube: Destroy").In(5.5).Seconds().Do(Object.Destroy());
}
The initial spike was me posting the project with visuals on reddit and various Discord channels, and commenting on Youtube videos. According to stats that actually didnât do that much, itâs mostly my Unity forum signature. Project still receives 1-3 stars per day.
I donât do much in Godot communities. Problem is: they donât allow user signatures unlike Unityâs forums. Creating a project thread is definitely planned but it requires frequent posts to keep it at the top of the list but I also donât want to be spammy.
I also have 158 Patreon subscribers which I email every 2-3 weeks.
I donât have any users yet. Itâs too early and even now user-facing API changes often.
This isnât helping beginners. The code complexity is so immense we severely dumb down the code to bad practices (ie all fields public, highly interdependent code, âthis.gameObject.transform.GetComponent
The easiest dismissal of the project is of course: why wouldnât I program specifically for the engine I am using? Why suffer the unavoidable overhead of a layered architecture (even if itâs <5%)? The need for switching engines is indeed rare, being able to move your code between engines is a niche pain (felt by contractors/agencies the most).
I think short term builds the long term. Beginners starting with LunyScript will keep it around, as it will continue to make the simple things simple, and it lets them wrap more complex code into simple executable blocks. They can easily slide into imperative programming, as a single line of block statement could just as well be written as part of a for loop - doing the same thing multiple times with variation.
Who is going to pay for it? Enthusiast individuals and educators, perhaps affluent learners. Perhaps strategic interests: security against another âruntime feeâ license change disaster, or growth and competition (ie make it easier for engine users to transition), and ecosystem growth (build a framework and sell it on several engine marketplaces).
Pivot: I wish to keep it open source, since a multi-engine SDK is a hard sell. Reasonably Iâd have to scale it down to Unity only (but keeping the agnostic layer) and sell it as an entry-level and designer-powerhouse which rivals visual scripting tools but without the visual clutter. It could lean into the behaviour observed in Godot. They cancelled their VisualScript because it had only a 0.5% userbase while users found GDScript easy enough, and it provided no high-level abstractions which make Blueprints so powerful.
My approach is different: I provide an engine-agnostic C# API with native engine integrations, whereas all other frameworks are engine-agnostic, they donât know nor care about the engineâs details (or they are 100% native, incompatible with any other engine, despite their general purpose). With LunyScript you can implement custom gameplay logic that is actually executing in the engine. LunyScript is the beginner-friendly, pattern-driven API. LunyEngine is the developer SDK with native engine abstractions on which LunyScript is built, and which other devs can utilize to write their own cross-engine frameworks. Itâs highly extensible and allows even for native-engine-only features ie one could write and provide executable blocks for Unityâs cloud services which would either only work in Unity, or which would utilize only Unityâs web cloud APIs to make it work in any engine.
Shippable in 4 weeks: I was thinking a simple Vampire Survivors, top-down but in 3d with Kenney.nl assets, crudely animated Minecraft-like characters, shooting projectiles, physics ragdolls on âdeathâ, plenty vfx/audio, minimalist UI. Most of the on-screen action is prefabs - the API doesnât expose VFX but instead a vfx prefab gets instantiated which plays the vfx and destroys itself afterwards. Ragdoll is a simple toggle switch too. This is part of the strength of LunyScript to be the âglueâ for the engine editor. Main issue I see here: Iâd have to do every asset twice. So Iâm not sure whether to do both engines and keep it simpler, or do one engine well.
Marketing: I have a website https://lunyscript.com with a narrative and visuals. Only architectural documentation yet. I still need to setup doxygen or similar, write getting started tutorials, record demo and tutorial videos. Itâs a lot ⌠and hard to say where time is best spent (probably videos, then getting started and example scripts).
Quick wins: I streamed a few time but I cannot write any ârealâ code while streaming. I could write LunyScript and play with that, once thereâs enough to show. Godot outreach: here I see a conflict since Godot heavily promotes GDScript but my solution is C#. Itâs currently still a better fit for Unity until Godot gets C# integrated into a single download and with web export. But I could leverage being among the first to teach Godot with C# (and in âthe easy wayâ).
I think I need the most help on what to do now (next two weeks) to make progress, stay motivated, and move towards Epic submission. Of course planning the whole path towards Megagrant submission (deadline: March 20th, today is Feb 17th). I should reserve one week for writing the application, even though I can repurpose a lot. Then in broader terms how to structure the three months thereafter (April, May, June).
I mostly need to start producing content, that means fleshing out and implementing the engine-native APIs that I need for a demo. Whether to focus only or primarily on Unity (10+ years experience). I should probably do both, but start with mechanics only (only primitives, no animations, vfx, ragdolls etc) and do the visual tweaking later which I could then polish in Unity first (it would look the best anyway).
Iâll need to keep track of the things I want to change, rather than refactor them right away.
Iâll read your PrototypeFund application to understand your previous pitch, then provide strategic guidance.Now let me provide strategic guidance based on everything youâve shared.
You asked about Unity/Godot bridges - there arenât really any doing what youâre doing. Closest comparisons:
Your approach is genuinely novel. The closest analog is Unrealâs Verse (Epicâs new language for UEFN/Fortnite) - which is declarative, beginner-friendly, and cross-platform. This is actually perfect positioning for Epic.
Goal: Playable game prototype that showcases LunyScriptâs power
What to build: Simple 3D arena shooter/survivor (your Vampire Survivors idea is good)
Deliverable: 2-3 minute video showing:
Goal: Package the demo for Epic reviewers
Create:
Goal: Submit a compelling Megagrant application
Application narrative shifts:
Key differentiators from PrototypeFund:
Ask amount strategy:
Based on typical Megagrant recipients:
â Strong signals:
â Weak signals:
Your pitch should emphasize:
Donât wait for Epic decision:
GitHub Sponsors (setup immediately)
Patreon refinement (you have 158 subs!)
Unity Asset Store (June-July)
Kickstarter (too early, maybe Q4 2025)
Consulting/Services (escape hatch)
The hard truth:
The opportunity:
What would make me fund this:
You can achieve 1-3 in the next month.
Daily rhythm:
End of week 1 target:
End of week 2 target:
Based on your request, hereâs what I propose we work on together:
A) Create a detailed 2-week development plan for the demo game B) Analyze your current codebase to identify whatâs missing for the demo C) Draft an Epic Megagrant application outline D) Something else?
What feels most valuable right now?
Weaknesses:
Competition context:
Before I gave you my application document, you mentioned âI know of some Unity/Godot bridge attempts but they're limitedâ .. what are they? Or do they overlap entirely with those I have in the application document?
Key: Every mechanic implemented in LunyScript shown side-by-side with Unity equivalent You mean I should re-implement the game with Unityâs native API simultaneously (or after the fact)? Thatâs quite the extra workload. How âkeyâ do you think this is? I could to a spot-check code comparison, and the rest extrapolates in someoneâs mind who is familiar with Unityâs API. I hope thatâll be enough. I could play it this way: the entire game is say 400 lines of LunyScript. Then I implement just as much as I can do in 400 lines with Unityâs API, and compare the results. Whatâs your take on this angle?
Should that demo feature a UI? Ie âchoose an updateâ on level-up, but Iâd rather follow the Brotato cycle of wave + shop.
Unreal integration:
Epic funds game tech True but Megagrant specifically lists âEducationâ, albeit this has been granted to curricula and Youtube tutorial channels, not the open source development channel. I view their âmake life better for the 3D communityâ statement as an open call. But true: focus should be more on the ecosystem, beginner-friendly (also: designer-/artist-friendly) as a side-effect.
I should probably highlight the pattern-based nature of my engine: coroutines, statemachine, behaviour tree patterns (i only have coroutines right now). Scripts execute once when the object comes alive (typically on scene load, at runtime i will aim for pooling rather than create/destroy cycles). I plan on runtime debugging tools: see (and modify) all global variables on screen. Select an object and inspect its local variables. Profile the script execution time (simple millisecond counter for now). Trace execution of blocks to see what runs and why. Placeholders for âmissingâ (typo?) assets, rather than hard fail (configurable). I should definitely highlight which of that is LunyEngine (cross-engine developer SDK) and what is LunyScript (block execution)
What âdeveloper pain pointsâ do you see that LunyEngine/LunyScript would, could, should address? Are there recurring patterns? Is what Iâm doing potentially disruptive, and if so, to what effect and would Epic benefit from that?
Diversification: Patreon + GH sponsors: already set up, currently $30 on Patreon and a one-time $10 on GH Patreon subscribers are âjoined for freeâ except for 2. Exclusive content is a good idea but perhaps only starting in 3+ months.
I absolutely agree on revenue stream NOW. Chances of getting a grant are slim.
You can achieve 1-3 in the next month. .. and maybe community validation. Interest is definitely there. So we should plan for this, despite lack of funding. Worst case: Project slows down next year, remains open source. Not-so-bad case: I take on a job which allows me to improve Luny either as part of the job, or by doing the job part-time only.
Start recording You suggest I should continuously record my desktop every day? If I do that, Iâll generate huge amounts of video. I should at least take notes of âoh that was interestingâ moments so I can delete the uninteresting recordings.
the daily rhythm fits mine very well I usually do 3-4h focused work. Then take 1-2h break. Do another 1-4h in the evening but ideally âlighterâ work as mental capacity is drained.
okay so I want to see what you got to say about my prompt here. then we can start working on the immediate next steps (architecture + roadmap only, items 3+4 are workflow related, no help needed) ask me again to do A and B at the end of your response
Honestly, I overstated it. When I said âI know of some Unity/Godot bridge attempts,â I was thinking of:
There is nothing like LunyScript. Your approach of providing engine-aware abstractions that execute native engine code is genuinely novel. The application document covers everything I was aware of.
Your â400 lines LunyScript vs 400 lines Unity MonoBehaviourâ angle is excellent and much smarter than parallel implementation.
Suggested approach:
Video narrative:
This is not critical-path for Epic, but itâs your strongest marketing asset. You could do this AFTER the application if time is tight.
Skip UI complexity. Brotatoâs wave + shop cycle requires:
Instead: Pure arcade survivor with score display only
Waves auto-progress based on timer/kill count. No choices, no shop, just escalating difficulty. You can add âweapon pickupsâ (touch to collect, auto-equip) if you want variety without UI.
My recommendation: Donât mention Unreal at all for now.
Better positioning:
Why:
What to emphasize instead:
Epicâs Megagrant evaluators will understand this is ambitious. They fund potential, not just current state.
Based on your description and code, here are the pain points youâre solving:
1. Tutorial Hell / Copy-Paste Programming
2. Engine Lock-In
3. Boilerplate Overload
4. Designer/Artist Programming Gap
5. Pattern Fragmentation
6. Debugging Friction
Yes, itâs disruptive in a ârising tide lifts all boatsâ way:
Disruption:
Epicâs Benefit:
Positioning for Epic: âLunyScript creates a cross-engine standard for gameplay code, reducing vendor lock-in and empowering developers. Epicâs funding would accelerate an open ecosystem that benefits all C# enginesâincluding future Unreal integration. This aligns with Epicâs demonstrated commitment to developer freedom (Godot funding, EGS revenue model) and creates goodwill across the game development community.â
This distinction is crucial for your pitch:
LunyEngine (Developer SDK):
LunyScript (Gameplay DSL):
This two-layer strategy is a strength:
Epic pitch should emphasize both:
Pattern-based architecture is a major strength:
On.Ready() syntax)This is your âBlueprints without visual clutterâ angle.
Debugging tools are table-stakes for designer tools:
These should be in your Epic application as âMilestonesâ:
Revenue NOW is critical. Hereâs the harsh timeline:
Parallel revenue streams (priority order):
Worst-case scenario acceptance: If Epic says no, youâll have 4-5 months left. With diversified revenue, you might stretch to 8-10 months. Thatâs enough time to:
Not-so-bad case (job with Luny overlap): Target companies:
Donât record continuously. Thatâs madness and storage hell.
Better approach:
Tools:
Weekly output:
This is sustainable and generates content without overhead.
You need 5-10 active users ASAP. Hereâs how:
Week 3-4 (post-demo):
Goal by Epic application deadline:
This is achievable in 4 weeks if you prioritize it.
Your plan is solid:
Key adjustments:
Next steps for immediate work:
Iâll analyze your current LunyScript API, identify gaps for the survivor demo, and create a detailed implementation roadmap (which blocks/APIs you need, in what order).
Iâll examine whatâs implemented vs whatâs needed for the demo, prioritize missing features, and flag technical debt to defer.
Which would you like me to start with: (A) Demo architecture plan, or (B) Codebase gap analysis?
Or should I do both in sequence (A first for planning, then B for validation)?