agentleFS
tim-osterhustim-osterhus/millrace-skills/skills/game-asset-pipeline-expansion/SKILL.md

game-asset-pipeline-expansion

Skill0 starsChanged 7 months agoApache-2.0View on GitHub

Guardrail for sprite-strip normalization, GLB / glTF prep, and browser-game asset verification.

The file

Show its rules and text

We split this skill into the rules it gives an agent

18 rules. About 0 are common, 17 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈17

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the matching reference file before touching the asset.

    line 31
  2. Distinctive:

    Run the sibling deterministic check on the prepared output, then render or inspect browser-facing preview evidence.

    line 33
  3. Distinctive:

    Keep the source-art boundary explicit in your notes so downstream agents know whether they are still editing the art or only preparing it …

    line 34
  4. Distinctive:

    Do not use this skill to generate or repaint source art.

    line 38
  5. Distinctive:

    Do not own runtime scene code, camera work, HUD work, or engine choice.

    line 39
  6. Distinctive:

    Do not collapse the sprite and 3D branches into one generic asset-management essay.

    line 40
  7. Very specific:

    Do not invent a new checker when the sibling visual-asset-pipeline/scripts/asset_check.py already covers deterministic validation.

    line 41
  8. Distinctive:

    Do not hide browser verification behind a local file check.

    line 42
  9. Distinctive:

    Keep the package portable: the scripts and references should be usable without private state.

    line 43
  10. Distinctive:

    Read the matching reference and choose the smallest safe prep step.

    line 63
  11. Distinctive:

    Run the package helper script for sprite normalization or preview generation, or use the named 3D export and optimization tools from the …

    line 64
  12. Distinctive:

    Run deterministic validation and inspect the preview or browser-facing evidence.

    line 65
  13. Distinctive:

    If the asset still fails loadability, scale, transparency, or frame-slot checks, repair the asset before claiming readiness.

    line 66
  14. Distinctive:

    Leave the source-art skill boundary intact so future work can still distinguish creation from prep.

    line 67
  15. Distinctive:

    Do not regenerate source art in this package.

    line 73
  16. Distinctive:

    Do not split sprite and 3D into separate skills unless the work itself needs that split.

    line 74
  17. Distinctive:

    Do not rely on preview polish alone when a deterministic check or browser run would expose the problem.

    line 75
  18. Distinctive:

    Start with the branch decision, then read only the matching reference.

    line 79

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill 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
SKILL.md

Game Asset Pipeline Expansion

Purpose

Use this skill when the asset already exists and the task is to make it browser-ready, not when the task is to create or repaint the art itself. It owns sprite-strip normalization and preview evidence for 2D assets, plus GLB / glTF preparation and deterministic verification for 3D assets. Keep the sprite details in references/sprite-normalization.md and the 3D details in references/web-3d-asset-prep.md; keep source-art generation in visual-asset-pipeline.

Quick Start

  1. Identify the branch: sprite-strip normalization or GLB / glTF prep.
  2. Read the matching reference file before touching the asset.
  3. Normalize or prep the asset with the package helper script or the external 3D export and optimization tools named in the reference.
  4. Run the sibling deterministic check on the prepared output, then render or inspect browser-facing preview evidence.
  5. Keep the source-art boundary explicit in your notes so downstream agents know whether they are still editing the art or only preparing it for ship.

Operating Constraints

  • Do not use this skill to generate or repaint source art.
  • Do not own runtime scene code, camera work, HUD work, or engine choice.
  • Do not collapse the sprite and 3D branches into one generic asset-management essay.
  • Do not invent a new checker when the sibling visual-asset-pipeline/scripts/asset_check.py already covers deterministic validation.
  • Do not hide browser verification behind a local file check.
  • Keep the package portable: the scripts and references should be usable without private state.

Inputs This Skill Expects

  • A prepared sprite strip, sprite source frame, or GLB / glTF asset that is already ready for prep work.
  • The intended game scale, frame count, output size, and anchor expectations for sprite work.
  • Texture-budget, collision, LOD, and loadability expectations for 3D work.
  • The target browser or browser-based runtime that will consume the asset.
  • Any existing preview, reference frame, or seed asset needed to lock scale or anchors.

Output Contract

  • Normalized sprite frames and a preview sheet or contact sheet for 2D work.
  • Browser-ready GLB / glTF output and a clear note about compression, collision, and LOD choices for 3D work.
  • Deterministic asset-check results from the sibling checker where applicable.
  • Honest blocker notes when an asset cannot be made ready without changing upstream art or runtime assumptions.

Procedure

  1. Classify the work as sprite or 3D and stay on that branch.
  2. Read the matching reference and choose the smallest safe prep step.
  3. Run the package helper script for sprite normalization or preview generation, or use the named 3D export and optimization tools from the reference.
  4. Run deterministic validation and inspect the preview or browser-facing evidence.
  5. If the asset still fails loadability, scale, transparency, or frame-slot checks, repair the asset before claiming readiness.
  6. Leave the source-art skill boundary intact so future work can still distinguish creation from prep.

Pitfalls And Gotchas

  • Rejected trope: a single generic asset-pipeline manual that tries to own art creation, runtime integration, and ship validation in one blob.
  • Better alternative: one narrow prep skill with a sprite reference, a 3D reference, and the sibling checker reused for deterministic checks.
  • Do not regenerate source art in this package.
  • Do not split sprite and 3D into separate skills unless the work itself needs that split.
  • Do not rely on preview polish alone when a deterministic check or browser run would expose the problem.

Progressive Disclosure

Start with the branch decision, then read only the matching reference. Pull in the other branch only when the task genuinely spans both sprite and 3D asset prep. Keep visual-asset-pipeline as the source-art sibling and this skill as the browser-game prep layer.

Verification Pattern

  • Confirm the sprite output has stable anchors, frame slots, and a preview sheet that exposes drift.
  • Confirm the 3D output is loadable, scaled correctly, and still within the texture and material budget the runtime can afford.
  • Confirm the sibling deterministic checker passes on the prepared asset where it applies.
  • Confirm the final advice ends at browser-game readiness, not at a generic file conversion step.

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.