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 nirholas/three.ws --skill mesh-forgegit clone --depth 1 https://github.com/nirholas/three.wsWrote 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/nirholas/three.ws/mesh-forge)<a href="https://agentmods.dev/skills/nirholas/three.ws/mesh-forge"><img src="https://agentmods.dev/badge/skills/nirholas/three.ws/mesh-forge.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00113 | $0.00743 |
| Opus 5 | $0.00056 | $0.00371 |
| Sonnet 5 | $0.00023 | $0.00149 |
| Haiku 4.5 | $0.00011 | $0.00074 |
Grade A, and why
mesh-forge 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 3d 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 — 46 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Mesh Forge — directed text/image/multi-view → 3D
The full-fidelity mesh lane: a chain of specialist models (IBM Granite prompt director → FLUX reference render → Microsoft TRELLIS / Tencent Hunyuan3D reconstruction), or direct reconstruction from your own photos.
Tool
mesh_forge (three.ws MCP server, @three-ws/mcp-server).
- Lane: PAID — $0.25 USDC per call over x402. Returns
PaymentRequiredif no payment payload is attached. - Modes:
- Text: Granite rewrites your prompt into an optimized 3D spec, FLUX renders a reference, TRELLIS/Hunyuan3D build the mesh.
- Image: pass
image_urlto reconstruct a single reference image directly (prompt-director + text-to-image stages skipped). - Multi-view: pass
image_urls(1–4 angles, e.g. front/back/left/right) and the backend fuses them — no hallucinated back of the object.
Inputs
| Param | Required | Notes |
|---|---|---|
prompt |
text mode | 3–1000 chars describing one object, e.g. "a worn leather armchair". Optional when an image is supplied (then used as guidance). |
image_url |
image mode | http(s) URL to a single reference image. |
image_urls |
multi-view | 1–4 http(s) URLs of the same object from different angles. Takes precedence over image_url. |
direct |
no | Run the Granite prompt-director stage (text mode). Default true. |
aspect_ratio |
no | Reference image aspect ratio for text mode: 1:1 (default), 4:3, 3:4, 16:9, 9:16. |
Outputs
glbUrl— the generated mesh.glb.viewerUrl—https://three.ws/viewer?src=<glbUrl>.- how many views were fused, which backend handled it, the directed prompt (text mode), and timing.
How to run
- Choose the mode from the user's input (text / single image / multi-view). Prefer multi-view when several angles of one object are available — it gives the cleanest geometry.
- Call
mesh_forge. OnPaymentRequired, surface the $0.25 USDC price and funding path rather than retrying. - Return
glbUrl+viewerUrl. FeedglbUrltoauto-rig(rig_mesh) for an animation-ready model.
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.
- 3d ago First seen · 46 lines · 113 tokens per session scan A 4234b60cedf6
mesh-forge is a skill published in the GitHub repository nirholas/three.ws (111 stars, last pushed yesterday), licensed Apache-2.0. It adds 113 tokens to every session and 743 once invoked, about $0.0006 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 2026-09-03.
Other skills, from other repositories
blender-mcp
Drive Blender via the catalog blender MCP, with bpy recipes.
blender-mcp
Control Blender directly from Gauss via socket connection to the blender-mcp addon. Create 3D objects, materials, animations, and run arbitrary Blender Python (bpy) code. Use when user wants to create or modify anything in Blender.
blender-mcp
Control Blender directly from Hermes via socket connection to the blender-mcp addon. Create 3D objects, materials, animations, and run arbitrary Blender Python (bpy) code. Use when user wants to create or modify anything in Blender.
blender-mcp
Control Blender directly from Hermes via socket connection to the blender-mcp addon. Create 3D objects, materials, animations, and run arbitrary Blender Python (bpy) code. Use when user wants to create or modify anything in Blender.
blender-mcp
Control Blender directly from Hermes via socket connection to the blender-mcp addon. Create 3D objects, materials, animations, and run arbitrary Blender Python (bpy) code. Use when user wants to create or modify anything in Blender.
blender-mcp
Drive Blender via the catalog blender MCP, with bpy recipes.