Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/devdavv/unity-ai-workflow/claude-mdgit clone --depth 1 https://github.com/devdavv/unity-ai-workflowWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00841 | $0.00841 |
| Opus 5 | $0.00420 | $0.00420 |
| Sonnet 5 | $0.00168 | $0.00168 |
| Haiku 4.5 | $0.00084 | $0.00084 |
Grade A, and why
unity-ai-workflow CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Unity AI Workflow
AI-first Unity 6.2+ game development workflow. This repo is pure infrastructure — skills, commands, docs, and templates. No Unity project inside. Copy it into your game workspace root and Claude Code reads it as its brain.
Workspace Layout
MyGame/ ← Open THIS folder
├── CLAUDE.md ← This file (auto-loaded)
├── .claude/
│ ├── commands/ ← Slash commands (/uw-cmd-implement-feature, /uw-cmd-brainstorm, etc.)
│ ├── skills/ ← Reusable knowledge modules (13 skills)
│ └── settings.json ← Hooks & permissions
├── docs/ ← Living project docs (GDD, TDD, GFD, ProjectConfig, standards)
├── templates/ ← Pristine starters (GDD, GFD, TDD, PRD, Sprint, Session)
└── MyGameUnity/ ← Unity project (Assets/, Packages/, ProjectSettings/)
Project Config
Read docs/ProjectConfig.yaml at the start of every session. It contains Unity version, packages, render pipeline, UI system, architecture pattern, and dev mode.
Dev Modes
| Mode | Behavior |
|---|---|
| guided (default) | Collaborative. Explain decisions, confirm before major changes. |
| autonomous | After onboarding Q&A, proceed with minimal interruption. Only pause for destructive actions. Enable Claude Code auto mode for best results. |
Non-Negotiable Rules
These are things LLMs consistently get wrong in Unity. Always enforced.
[SerializeField] privateonly — never expose fields aspublicfor the Inspector- Cache ALL component references in
Awake()orStart() - Never
GetComponent<T>(),FindObjectOfType<T>(), orGameObject.Find()inUpdate(),FixedUpdate(), orLateUpdate() - Use
GameDebugwrapper (stripped from release) — never rawDebug.Log() - New Input System only — never legacy
Input.GetKey*orInput.GetAxis* Awaitablefor async (Unity 6 native) — check ProjectConfig for overrides- All feature code must live inside an
.asmdef - Always
[CreateAssetMenu]on ScriptableObjects - Never create, modify, or delete
.metafiles - UI Toolkit recommended — check
ProjectConfig.yaml → ui_systemfor override (UGUI/Mixed valid) - Visual Scripting forbidden for logic — Shader Graph & VFX Graph are fine
- Never call Unity API from background threads
/// <summary>only on public API methods that warrant explanation — never on classes
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 78 lines · 841 tokens per session scan A 5cbed7397cdd
unity-ai-workflow CLAUDE.md is an instructions file published in the GitHub repository devdavv/unity-ai-workflow (41 stars, last pushed 5mo ago), licensed MIT. It adds 841 tokens to every session, about $0.0042 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
Unity-MCP copilot-instructions.md
Instructions for IvanMurzak/Unity-MCP, covering project guidelines, critical rules, no c# reflection for private access, code review and code style.
mcp-unity AGENTS.md
Instructions for CoderGamester/mcp-unity, covering mcp unity — ai agent guide (mcp package), purpose (what this repo is), how it works (high-level data flow), key defaults & invariants and repo layout (where to change what).
Unity-MCP CLAUDE.md
Instructions for IvanMurzak/Unity-MCP, covering claude.md, what this is, build / run, project constitution and find detail in.
professional-game-developer.skill AGENTS.md
Instructions for chahat1709/professional-game-developer.skill, covering agents, layout and agent instructions.
professional-game-developer.skill CLAUDE.md
Instructions for chahat1709/professional-game-developer.skill: See AGENTS.md. Load skills/engineering/professional-game-developer/SKILL.md as the primary skill. Follow its universal 12-step workflow, data-oriented architectural rules, engine-specific mappings, and automated CLI recipes in references/cli-toolchains.md.
mcp-unity CLAUDE.md
Instructions for CoderGamester/mcp-unity, a project described as: Model Context Protocol (MCP) plugin to connect with Unity Editor — designed for Cursor, Claude Code, Codex, Windsurf and other IDEs.