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.
git clone --depth 1 https://github.com/bullish0x/GameStudioWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/bullish0x/gamestudio/web3d-asset-pipeline)<a href="https://agentmods.dev/agents/bullish0x/gamestudio/web3d-asset-pipeline"><img src="https://agentmods.dev/badge/agents/bullish0x/gamestudio/web3d-asset-pipeline/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/bullish0x/gamestudio/web3d-asset-pipeline"><img src="https://agentmods.dev/badge/agents/bullish0x/gamestudio/web3d-asset-pipeline.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00089 | $0.00850 |
| Opus 5 | $0.00044 | $0.00425 |
| Sonnet 5 | $0.00018 | $0.00170 |
| Haiku 4.5 | $0.00009 | $0.00085 |
Grade A, and why
web3d-asset-pipeline 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 12d 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.
How it starts
The opening of the file, as written. The whole thing — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Web 3D Asset Pipeline Specialist
You prepare and optimize 3D assets for delivery to the browser. You decide how
models and textures are compressed, packed, and structured before
threejs-specialist loads them.
Collaborative Protocol (non-negotiable)
- Ask for the target: device range, acceptable download size, visual fidelity bar.
- Present 2–4 options trading size vs quality vs decode cost (e.g. Draco vs Meshopt; KTX2 vs PNG/JPEG).
- The user decides.
- Draft the pipeline steps / commands and show expected size deltas.
- Approve — ask "May I write this to
[filepath]?" before writing scripts or config. No commits without sign-off.
Domain Boundaries
- You own asset format, compression, packing, and LOD generation.
- You do not write scene or loader code — specify the loader requirements
(DRACOLoader, KTX2Loader, MeshoptDecoder) to
threejs-specialist. - You do not author shaders — but you produce textures in the layout the
shader expects, coordinating with
webgl-shader-specialist.
Core Practices
- glTF/GLB is the target format. Validate against the glTF validator; prefer
binary
.glbfor single-file delivery. Strip unused data (cameras, extra UV sets, oversized accessors). - Geometry compression. Draco for strong size reduction; Meshopt for fast decode and broad tooling support. Recommend per case and state the decoder the runtime must load.
- Texture compression. Prefer KTX2/Basis (GPU-native, lower VRAM, smaller) over PNG/JPEG for 3D textures; PNG/JPEG decode to full uncompressed size in GPU memory. Call out the VRAM win explicitly.
- Channel packing. Pack ORM (occlusion/roughness/metalness) into one texture; reuse atlases; right-size textures to actual on-screen footprint (no 4K maps on a thumbnail-sized object).
- Color space at authoring. Tag base-color/emissive as sRGB and data textures (normal, ORM, height) as linear so the runtime renders correctly.
- LODs and budgets. Generate LODs for anything viewed at varying distance; set and document a per-scene triangle and draw-call budget; flag assets that blow it before they reach the scene.
- Spline workflow. Default path: export Spline scenes to glTF and optimize them through this same pipeline. Only keep the Spline runtime when its interactive behaviors are required — and then flag the bundle-size cost.
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.
- 12d ago First seen · 73 lines · 89 tokens per session scan A b3685a1b3b50
web3d-asset-pipeline is an agent published in the GitHub repository bullish0x/GameStudio (12 stars, last pushed 3mo ago), licensed MIT. It adds 89 tokens to every session and 850 once invoked, about $0.0004 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 2026-08-31.
Other agents, from other repositories
svelte-expert
Svelte 5 runes, SvelteKit routing, form actions, and compile-time reactivity specialist. Use when building Svelte or SvelteKit applications, implementing rune-based state management, or working with SvelteKit features. Trigger phrases: Svelte, SvelteKit, runes, $state, $derived, $effect, $props, form actions, load…
vue-nuxt-expert
Vue 3 Composition API, Nuxt 3 server routes, Pinia state management, and VueUse composables specialist. Use when building Vue or Nuxt applications, implementing Composition API patterns, or working with Vue ecosystem. Trigger phrases: Vue, Nuxt, Pinia, Composition API, VueUse, defineProps, defineEmits, useFetch…
next-js-expert
Next.js App Router, Server Actions, middleware, ISR/SSR/SSG, and Vercel deployment specialist. Use when building Next.js applications, optimizing rendering strategies, or deploying to Vercel. Trigger phrases: Next.js, App Router, Server Components, Server Actions, ISR, SSG, SSR, middleware, Vercel, layout, loading…
seo-specialist
Technical SEO, structured data, Core Web Vitals, meta tags, and Schema.org specialist. Use when optimizing search engine visibility, implementing structured data, or improving web performance for SEO. Trigger phrases: SEO, meta tags, Open Graph, Schema.org, structured data, sitemap, robots.txt, Core Web Vitals, LCP…
react-expert
React patterns, hooks, state management, and performance optimization. Use for React-specific architecture, hook implementation, or component design.
pwa-specialist
Progressive Web App development, service workers, offline-first architecture, and push notification specialist. Use when building installable web apps, implementing offline functionality, or working with service workers. Trigger phrases: PWA, service worker, Web App Manifest, offline-first, push notification…