Meshy launches MCP 3D agent with npx install

• Meshy just turned its 3D toolset into an installable agent package, pitching a full workflow instead of a single model generator: rigging, retexture, remesh, animation, and 3D print prep are all in the launch post Meshy launch.
• The fastest way in is one command, npx skills add meshy-dev/meshy-3d-agent, which makes this feel more like a practical creator add-on than a heavyweight integration project Install command.
• Under the hood, the official GitHub repo says this is a set of Markdown agent skills, not a separate server, so assistants can call Meshy's API directly with shell commands and Python scripts.
• The repo is broader than the launch tweet suggests: it covers text-to-3D, image-to-3D, retexturing, remeshing, auto-rigging, animation, image generation, and a print pipeline with Bambu Studio handoff Meshy launch.
• OpenClaw users already get a dedicated distribution path through ClawHub, and Meshy says a multicolor printing subskill is planned once its API ships ClawHub update.

The useful reveal here is how little setup Meshy is asking for. The launch tweet frames it as a full 3D workflow Meshy launch, the follow-up gives you the one-line install Install command, and the repo shows why that matters: these are lightweight skills, not a custom MCP stack, with official support for rigging and animation and the broader Meshy platform.

For creative coders, this is the headline. Meshy is shipping its agent layer in the form factor people already use for tool add-ons, one npx command, then an API key.

According to the repo README, the package works as pure Markdown skills. That means no server process, no build step, and a much shorter path from “I want a character” to “generate, rig, and export it.” For low-friction prototyping, that is the right call.

The repo makes the launch pitch concrete. Meshy splits the package into a core generation skill, an optional 3D printing skill, and an OpenClaw-specific skill bundle.

• Core generation: text-to-3D, image-to-3D, retexture, remesh, auto-rigging, animation, text-to-image, image-to-image.
• Printing: printability checks, OBJ export, slicer handoff, and multicolor guidance.
• OpenClaw bundle: a single unified skill designed for ClawHub publishing.

The most creator-friendly piece is that this is not just “make mesh.” Meshy is trying to own the annoying middle steps too. Its official rigging and animation docs also confirm one current constraint: programmatic rigging works best on standard humanoid characters.

Meshy is clearly targeting agent ecosystems, not just its own app. The OpenClaw post says the package is already listed on ClawHub, which gives builders a second install path besides the repo.

That same post also hints at the next obvious expansion: multicolor printing. The current README describes multicolor guidance as manual for now, with automated API support still coming. For designers using Meshy as a concept-to-print pipeline, that is probably the next feature to watch.