unity-performance

A Unity game-performance guide covering how to measure and reduce CPU, graphics-card, and memory costs.

In plain words
What is it for?
Use it for profiling, frame and memory analysis, draw-call and batching reviews, object pooling, and optimization of frequently repeated game code.
Why use it?
It helps developers find the actual cause of slow frame rates, excessive memory use, or other performance problems before changing code.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/thearcforge/uniclaude/unity-performance
Any agent
npx skills add TheArcForge/UniClaude --skill unity-performance
Clone the repo
git clone --depth 1 https://github.com/TheArcForge/UniClaude

Made for: Claude Code, Codex.

Per session 33 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,192 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What 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.

ModelPer sessionOnce invoked
Fable 5 $0.00033 $0.01192
Opus 5 $0.00016 $0.00596
Sonnet 5 $0.00007 $0.00238
Haiku 4.5 $0.00003 $0.00119

Measured 2d ago against content hash 21a4e74fa9bb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

unity-performance 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 2d ago.

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.

Sidecar~/skills/unity-performance/SKILL.md · 136 lines

How it starts

The opening of the file, as written. The whole thing — 136 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Unity Performance — Full Decision Framework

Use this when unity-architect's condensed performance section needs more depth.

Before Optimizing

Profile first. Never optimize based on assumptions.

Tools:

  • Unity Profiler (CPU, GPU, Memory modules)
  • Frame Debugger (draw calls, batching, overdraw)
  • Memory Profiler package (heap snapshots, native allocations)
  • project_get_console_log — check for warning spam (often a hidden perf cost)

Rule: Identify the bottleneck, then optimize that specific thing. Optimizing the wrong thing is wasted effort.

CPU Cost Model

Expensive (avoid in hot paths):

  • GetComponent() — reflection-based lookup. Cache in Awake().
  • GameObject.Find() / FindObjectOfType() — linear scene scan. Never in Update.
  • string concatenation — allocates new string each time. Use StringBuilder.
  • LINQ in Update — allocates enumerators and closures.
  • Instantiate/Destroy — heavyweight. Pool for frequent use.
  • Physics.Raycast (many per frame) — use NonAlloc variants, limit ray length.
  • SendMessage / BroadcastMessage — reflection-based. Use direct calls or events.

Cheap:

  • transform.position (cached by engine)
  • Comparing tags (CompareTag, not ==)
  • Null checks on Unity objects (but NOT ?? or ?. — those bypass Unity's null check)
  • Static method calls, struct operations
  • Fixed-size arrays, pre-allocated lists

GPU Cost Model

Draw calls (CPU→GPU overhead per rendered object):

  • Each unique material = at least one draw call.
  • Reduce by: batching (static/dynamic/GPU instancing/SRP batcher), atlasing textures.
  • Target: < 100 on mobile, < 2000 on console, < 5000 on PC.

Fill rate (per-pixel cost):

  • Transparent objects drawn back-to-front, no early-Z rejection.
  • Overdraw: overlapping transparent objects multiply pixel cost.
  • Reduce by: smaller particles, opaque where possible, LOD.

Shader complexity:

  • Per-vertex operations are cheap (scale with mesh complexity).
  • Per-pixel operations are expensive (scale with screen resolution).
  • Texture samples per pixel: each costs bandwidth. Combine maps.

Read the full file on GitHub · 136 lines

Changes

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.

  1. 2d ago First seen · 136 lines · 33 tokens per session scan A 21a4e74fa9bb

Subscribe to this mod's changes

unity-performance is a skill published in the GitHub repository TheArcForge/UniClaude (51 stars, last pushed 3mo ago), licensed MIT. It adds 33 tokens to every session and 1,192 once invoked, about $0.0002 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.