agentsclimarketplace

Asset optimization

Skill arjun988/blender-skills/.cursor/skills/asset-optimization

94 specialist Blender skills for Cursor, Claude Code, Kiro & Codex -> MCP-powered production pipeline from blockout to engine export.

Install
npx -y skills add arjun988/blender-skills --skill asset-optimization

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

What its author says it does

Copied from the file, not written here

Validates and optimizes Blender assets for polycount, topology, UV efficiency, material count, naming, collections, and game-ready performance. Use before export on all production assets via MCP.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

3.7 KB, 774 tokens by cl100k_base, as published. Nobody here has run it

Asset Optimization

Enforce budgets. Clean geometry. Game-ready validation gate before export.

When to Use

  • Before every export
  • After modeling complete
  • LOD generation
  • Scene-wide performance audit
  • Pre-delivery validation

Workflow

Audit → Identify Issues → Fix → Re-Audit → LOD Generation → Validation Report

Audit Checklist

Run via MCP scene queries where possible:

Geometry

  • Triangle count vs budget (../references/polycount-budgets.md)
  • Duplicate vertices merged
  • Loose geometry removed
  • Non-manifold fixed
  • Interior faces deleted
  • Applied transforms (scale 1,1,1)

UVs

  • Pack efficiency > 75%
  • Texel density consistent
  • No unintended overlaps

Materials

  • Count within budget
  • No unused slots
  • Names follow MAT_ convention

Organization

  • All objects named SM_ prefix
  • Collections organized per convention
  • Collision meshes in COL_Collision

Optimization Techniques

IssueFix
Over budgetRemove unseen geo, decimate, bake detail
High triangle count on flat surfacesReduce subdiv/bevel segments
Many draw callsMerge materials; atlas textures
Repeated geometryConvert to instances
Heavy modifiersApply and merge strategically

LOD Generation

LOD0: Full detail (hero)
LOD1: 50% tris — remove micro-bevels, small greebles
LOD2: 25% tris — decimate modifier; preserve silhouette
LOD3: 10% tris — impostor/card or simple proxy

Naming: SM_Asset_LOD0, SM_Asset_LOD1, etc.

Use Decimate modifier:

  • Planar mode for hard surface
  • Collapse mode for organic
  • Always verify silhouette after decimate

Modifier Cleanup

Before export:

  1. Apply stack OR export with modifiers per engine support
  2. Remove unused modifiers
  3. Collapse mirror/array if instances not needed
  4. Document which modifiers remain live

Scene Audit Report Template

## Optimization Report: [Asset/Scene]

**Total Triangles:** [count] (budget: [target])
**Draw Calls (materials):** [count]
**Objects:** [count]
**Issues Fixed:** [list]

### Per-Object Breakdown
| Object | Tris | Materials | Status |
|--------|------|-----------|--------|
| SM_... | ... | ... | PASS/FAIL |

MCP Integration

  1. Query polycount per object via MCP
  2. Query material count via MCP
  3. Run cleanup operations via MCP
  4. Generate LOD duplicates via MCP
  5. Validate naming conventions via MCP scene tree

Constraints

MUST DO

  • Run full validation before export
  • Document polycount in production brief
  • Create LODs for hero assets in games
  • Fix all non-manifold before export
  • Output validation report

MUST NOT DO

  • Export without budget check
  • Decimate deforming meshes without testing
  • Remove collision proxies
  • Skip naming convention validation
  • Optimize hero face/hands below animation needs

Reference Guide

TopicReferenceLoad When
Full validation../references/validation-checklist.mdFinal gate
Polycount budgets../references/polycount-budgets.mdBudget setting
LOD strategiesreferences/lod-strategies.mdLOD creation

Gives 0 of the 12 instructions most performance cost skills give in 774 tokens

Counted across 803 of the 1,058 authors here whose files we hold, read 2026-08-07

  • keep skill files under 500 linesin 82 of 803, across 16 files
  • use imperative form in instructionsin 80 of 803, across 9 files
  • draft assertions while test runs are in progressin 75 of 803, across 9 files
  • create two to three realistic test promptsin 74 of 803, across 9 files
  • write skill descriptions to be pushyin 72 of 803, across 7 files
  • save test cases to evals jsonin 72 of 803, across 6 files
  • ask questions about edge cases and input formatsin 72 of 803, across 7 files
  • save timing data immediately when runs completein 70 of 803, across 5 files
  • include all trigger conditions in the skill descriptionin 69 of 803, across 3 files
  • launch all test runs in a single turnin 69 of 803, across 3 files
  • capture intent before writing a skillin 67 of 803, across 1 file
  • import directly instead of barrel filesin 52 of 803, across 15 files

Said here and by no other author read

  • Run full validation before export
  • Fix all non-manifold geometry before export
  • Create LODs for hero game assets
  • Merge duplicate vertices
  • Remove unused material slots
  • Prefix static mesh names with SM_

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.