Claude Code skills that build complete Godot games
Details
- External ID
- 47400868
- Source
- HN
- Company
- —
- Product
- skills
- Website domain
- github.com
- Launched
- March 16, 2026
- Cohort
- —
- Upvotes
- 337
- Upvotes percentile
- 0.9846248462484625
- Tags
- —
- Fetched at
- Sept. 7, 2026, 9:26 p.m.
- Updated at
- Sept. 7, 2026, 9:26 p.m.
Description
I’ve been working on this for about a year through four major rewrites. Godogen is a pipeline that takes a text prompt, designs the architecture, generates 2D/3D assets, writes the GDScript, and tests it visually. The output is a complete, playable Godot 4 project.Getting LLMs to reliably generate functional games required solving three specific engineering bottlenecks:1. The Training Data Scarcity: LLMs barely know GDScript. It has ~850 classes and a Python-like syntax that will happily let a model hallucinate Python idioms that fail to compile. To fix this, I built a custom reference system: a hand-written language spec, full API docs converted from Godot's XML source, and a quirks database for engine behaviors you can't learn from docs alone. Because 850 classes blow up the context window, the agent lazy-loads only the specific APIs it needs at runtime.2. The Build-Time vs. Runtime State: Scenes are generated by headless scripts that build the node graph in memory and serialize it to .tscn files. This avoids the fragility of hand-editing Godot's serialization format. But it means certain engine features (like `@onready` or signal connections) aren't available at build time—they only exist when the game actually runs. Teaching the model which APIs are available at which phase — and that every node needs its owner set correctly or it silently vanishes on save — took careful prompting but paid off.3. The Evaluation Loop: A coding agent is inherently biased toward its own output. To stop it from cheating, a separate Gemini Flash agent acts as visual QA. It sees only the rendered screenshots from the running engine—no code—and compares them against a generated reference image. It catches the visual bugs text analysis misses: z-fighting, floating objects, physics explosions, and grid-like placements that should be organic.Architecturally, it runs as two Claude Code skills: an orchestrator that plans the pipeline, and a task executor that implements each piece in a `context: fork` window so mistakes and state don't accumulate.Everything is open source: https://github.com/htdt/godogenDemo video (real games, not cherry-picked screenshots): https://youtu.be/eUz19GROIpYBlog post with the full story (all the wrong turns) coming soon. Happy to answer questions.
Enrichment
- Theme
- Claude integrations and coding agents
- Vertical
- Horizontal
- Function
- Dev tools
- Audience
- Developer
- AI stance
- AI feature
- Project type
- Hobby / open-source project
- Normalized one-liner
- ai-assisted godot game development
- Manually corrected
- False
Could you build this?
Partial While prompting LLMs to emit GDScript can be vibe-coded, reliably generating complete playable games requires deep expertise in Godot engine architecture, asset pipelines, and automated headless visual verification.
What it would actually take: The architecture requires a Python or Node.js orchestrator interfacing with LLM APIs, multi-modal asset generators, and a headless Godot 4 engine runner. The hardest part is deterministic GDScript generation, AST validation, scene tree serialization (.tscn), and computer-vision-based automated gameplay testing to catch runtime failures. This requires deep domain knowledge of game engine internals and game mechanics design.
Discussion
20 comments analyzed.
Competitors mentioned: Claude Code (used alongside this product), Manual game development in Godot
Concerns raised: GDScript's unique syntax trips up standard LLMs, Art/animation generation quality inconsistent (2D sprites messy), Messaging about "builds complete games" misleading to users, Performance-critical code still requires manual optimization, Testing/evaluation loop difficult with complex games
Feature requests: Integrated art creation pipeline documentation, Better 2D animation generation approach, Improved evaluation/testing workflow for modifications, Direct Claude Code integration into Godot editor
Competitors
Other products that read as similar to this one — 305 launches clear the similarity bar, closest 8 shown.
Attention rank: #8 of 306 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).
Launched 135 days after the earliest competitor.
- I made a Mario Galaxy game with Claude Code and Three.js in 53 days · hn · 2026-04-01 · 7 upvotes · similarity 0.52
- Godot MCP Studio · ph · 2026-09-18 · 2 upvotes · similarity 0.48
- Vekmio · ph · 2026-09-29 · 1 upvotes · similarity 0.48
- MakeSkill · hn · 2025-11-27 · 9 upvotes · similarity 0.47
- Lessons learned from running Claude Code swarms at scale · hn · 2026-06-05 · 10 upvotes · similarity 0.47
- Godot and Rust based multiplexer (terminal panes and more) · hn · 2026-09-11 · 97 upvotes · similarity 0.44
- Forge-GPU · hn · 2026-02-27 · 5 upvotes · similarity 0.44
- Prawduct, a product development framework for Claude Code · hn · 2026-03-31 · 6 upvotes · similarity 0.44
Other launches for this product
- skills
- Hivemind turns agent traces into skills and shares with your team
- skills
- adeskills
- skills
- skills
- Ratel, give agents unlimited tools and skills without context bloat
- Ingot, evidence-gated optimization and version control for agent skills
- Colin, a context engine for keeping agent skills fresh
- Distributing AI agent skills via NPM
- Open-source skills for end-to-end video ad campaigns
- TLA+ Workbench skill for coding agents (compat. with Vercel skills CLI)
Same idea, different domain
Nobody's really built a dev tools tool for Sales yet.