# 3d-cog

> Use this tool when you need to process imperfect images for 3D reconstruction or analysis, as it can handle noisy or distorted inputs to generate accurate 3D models. It solves problems related to image processing and 3D modeling, such as reconstructing objects from blurry or low-resolution images. The 3d-cog tool takes in 2D images and outputs 3D models, making it a valuable asset in various fields like computer vision and robotics.

Canonical page: https://skillsregistry.net/skills/nitishgargiitd-3d-cog  
JSON: https://api.skillsregistry.net/v1/skills/nitishgargiitd-3d-cog

## Description

Other tools need perfect images.

## Trust

- **Trust score (0–1):** 0.60
- **Verification tier:** unverified

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** instructions
- **Runtime environment:** llm
- **Category:** media
- **Updated:** 2026-05-16

## Source

- **Source listing:** [ClawHub](https://clawskills.sh/skills/nitishgargiitd-3d-cog)

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "nitishgargiitd-3d-cog"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/nitishgargiitd-3d-cog` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/nitishgargiitd-3d-cog/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt