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 skills add JimCline/hy3d-mcp --skill create-3d-modelgit clone --depth 1 https://github.com/JimCline/hy3d-mcpWrote 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/skills/jimcline/hy3d-mcp/create-3d-model)<a href="https://agentmods.dev/skills/jimcline/hy3d-mcp/create-3d-model"><img src="https://agentmods.dev/badge/skills/jimcline/hy3d-mcp/create-3d-model/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/skills/jimcline/hy3d-mcp/create-3d-model"><img src="https://agentmods.dev/badge/skills/jimcline/hy3d-mcp/create-3d-model.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 | $0.00066 | $0.03056 |
| Opus 5 | $0.00033 | $0.01528 |
| Sonnet 5 | $0.00013 | $0.00611 |
| Haiku 4.5 | $0.00007 | $0.00306 |
Grade A, and why
create-3d-model 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 8d 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 — 242 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Create a 3D model from a prompt or image
You have the hy3d-gen MCP server (bundled with this plugin): local
image-to-3D via Hunyuan3D-MLX. The pipeline is concept image → RGBA cutout
→ shape + PBR paint → GLB. Your job is to get the user from "I want a 3D
model of X" to a GLB file path, with a preview when possible.
First use in a session
Call server_status once before the first generation. Don't re-check on
later calls.
If any check is ok: false, the engine behind this server isn't set up
yet. Call setup_engine — it defaults to a dry run, changes nothing, and
returns the plan. Show that plan to the user, costs included (a ~4 minute
swift build, a ~12GB weight download), and only call it again with
confirm=true once they agree. Phases are idempotent, so a re-run after
a failure resumes rather than restarting.
If they would rather do it by hand, relay the failing check's fix text
verbatim instead — every failure carries its exact remedy.
Step 1 — get a concept image
If the user supplied an image, use it directly and go to Step 2.
If the user gave a text prompt, generate a concept image with whatever image-generation tool is available in the session (e.g. a Gemini or other image-gen MCP). If none is available, ask the user for an image — do not try to proceed without one.
The concept image makes or breaks the model, and it is the single biggest
lever on output quality — bigger than any generator knob. Measured on one
subject at identical seed and defaults, a clean evenly-lit input produced
31% more geometry (101k vs 77k verts) and visibly crisper panel and
edge detail than the painted concept art of the same object. Spend effort
here before reaching for octree.
Compose the image prompt from the user's description plus ALL of these:
- single object, whole object in frame, roughly centered
- three-quarter view (shows front and side; best geometry recovery)
- plain, uniform background — nothing else in frame, and a flat single tone rather than a gradient or vignette. The cutout keys on corner colour, so hard figure/ground separation matters. Light gray is the default. When the subject has no white or near-white parts, pure white with product-cutout framing ("isolated on seamless white, catalog product cutout") is worth trying — it came back clean once where two gray-background attempts kept a contact shadow through increasingly emphatic negative prompting. That is a single sample, and it changed the background colour and the framing language together, so which part did the work is unknown.
- even, soft, neutral studio lighting. Ask for "soft even studio lighting, neutral white". Avoid dramatic, moody, rim-lit, golden-hour or single-hard-key looks: baked-in directional shading and blown highlights are reconstructed as surface relief that isn't there. But never request "flat lighting", "unlit", or "albedo style" either — the generator de-lights internally, and pre-flattened input bakes pale and featureless. Even and soft, not absent.
- no drop shadow, no contact shadow — shadows under the object are reconstructed as literal geometry. Say "floating, no shadow" in the prompt.
- no text, watermark, or frame
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.
- 8d ago First seen · 242 lines · 66 tokens per session scan A 52c3dfc57af8
create-3d-model is a skill published in the GitHub repository JimCline/hy3d-mcp (0 stars, last pushed 1mo ago), licensed MIT. It adds 66 tokens to every session and 3,056 once invoked, about $0.0003 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 skills, from other repositories
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
general-video
Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated…
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
diagnostic-stem-delivery
Audio production with diagnostic analysis, timecode parsing from documents, and verified export workflow.
chengfeng-check-updates
An environment manager for a video-editing system. It checks whether its skills and runtime—the software needed to run them—are installed and compatible.