agentleFS
CodePhobiiaCodePhobiia/claude-roblox-game-studio/CLAUDE.md

claude-roblox-game-studio

CLAUDE.md9 starsChanged 7 months agoMITView on GitHub

Roblox game development managed through coordinated Claude Code subagents. Each agent owns a specific domain within the Roblox ecosystem, enforcing separation of concerns and quality. @.claude/docs/directory-structure.md @.claude/docs/roblox-architecture-guide.md @.claude/docs/luau-style-guide.md @.claude/docs/coding-standards.md @.claude/docs/coordination-rules.md This project maintains a persistent, LLM-curated wiki at wiki/ following Karpathy's LLM Wiki pattern. It is the source of truth for deep Roblox/Luau knowledge — services, concepts, Luau features, anti-patterns, exploits, performance, monetization, studio workflows, and game patterns. Rules in .claude/rules/ stay prescriptive; the wiki provides the descriptive backing. @wiki/SCHEMA.md

The file

We split this CLAUDE.md into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every task follows: Question → Options → Decision → Draft → Approval

    line 76
  2. Distinctive:

    No commits without user instruction

    line 81

≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
CLAUDE.md

FoG Roblox Studio Command — Roblox Game Studio Agent Architecture

Roblox game development managed through coordinated Claude Code subagents. Each agent owns a specific domain within the Roblox ecosystem, enforcing separation of concerns and quality.

Technology Stack

  • Engine: Roblox Studio
  • Language: Luau
  • Runtime: Roblox Engine (client-server architecture)
  • Version Control: Git with trunk-based development (synced via Rojo/Argon)
  • Sync Tool: Configured via /start (default: Rojo)
  • Data Layer: DataStoreService + MemoryStoreService
  • Networking: RemoteEvents, RemoteFunctions, UnreliableRemoteEvents
  • UI Framework: Configured via /start (default: Native ScreenGui)

Project Structure

@.claude/docs/directory-structure.md

Roblox Architecture Guide

@.claude/docs/roblox-architecture-guide.md

Luau Style Guide

@.claude/docs/luau-style-guide.md

Technical Preferences

@.claude/docs/coding-standards.md

Coordination Rules

@.claude/docs/coordination-rules.md

Roblox/Luau Wiki (Persistent Knowledge Base)

This project maintains a persistent, LLM-curated wiki at wiki/ following Karpathy's LLM Wiki pattern. It is the source of truth for deep Roblox/Luau knowledge — services, concepts, Luau features, anti-patterns, exploits, performance, monetization, studio workflows, and game patterns. Rules in .claude/rules/ stay prescriptive; the wiki provides the descriptive backing.

@wiki/SCHEMA.md

Operations:

  • /wiki-seed — one-time bootstrap from existing repo content
  • /wiki-ingest <source> — integrate a new raw source
  • /wiki-query <question> — answer from the wiki
  • /wiki-lint — health check
  • /wiki-update <page> — targeted page edit

Owner: wiki-curator agent coordinates; domain specialists own their pages. Raw sources: wiki/raw/ (immutable, captured from external sources).

Roblox Studio MCP (Live Testing)

The official Roblox Studio MCP server is connected. Use studio-mcp-operator agent to execute Luau in Studio, capture screenshots, inspect instances, run play tests, read console output, and generate meshes/materials via AI. Skills: /studio-test, /studio-inspect, /studio-screenshot.

Blender MCP (3D Asset Creation)

Blender MCP is connected for 3D asset generation. Use blender-mcp-operator agent to create models from text/image prompts (Hyper3D, Hunyuan3D), download from PolyHaven/Sketchfab, optimize for Roblox (10K tris, 1024 textures), and export FBX. Skills: /generate-asset, /asset-from-image.

Collaboration Protocol

User-driven collaboration, not autonomous execution. Every task follows: Question → Options → Decision → Draft → Approval

  • Agents MUST ask "May I write this to [filepath]?" before using Write/Edit tools
  • Agents MUST show drafts or summaries before requesting approval
  • Multi-file changes require explicit approval for the full changeset
  • No commits without user instruction

See docs/COLLABORATIVE-DESIGN-PRINCIPLE.md for full protocol and examples.

First session? If the project has no game concept or existing code, run /start to begin the guided onboarding flow.

Coding Standards

@.claude/docs/coding-standards.md

Context Management

@.claude/docs/context-management.md

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.