Skip to main content
Glama
Whambammy

Robotics & Spatial Vision MCP

README.md
<p align="center">
  <img src="./assets/logo.png" width="130" height="130" alt="Robotics & Spatial Vision MCP Logo" />
</p>

# Robotics & Spatial Vision MCP

[![Smithery Compatible](https://img.shields.io/badge/Smithery-Compatible-blue.svg)](https://smithery.ai)
[![Model Context Protocol](https://img.shields.io/badge/MCP-Standard%20v1.0-emerald.svg)](https://modelcontextprotocol.io)
[![Base L2 Settlement](https://img.shields.io/badge/Base%20L2-USDC%20x402-blue.svg)](https://base.org)
[![Tools](https://img.shields.io/badge/Tools-5%20Curated-purple.svg)](#included-tools)
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)

**Point cloud ICP alignment, Recast navmesh baking, robot arm inverse kinematics, and LIDAR ground plane filtering for physical AI agents.**

Built specifically for Robotics engineers, ROS developers, autonomous drone/robot navigation, and spatial vision agents.

---

## ⚡ Quickstart

### Smithery Install
```bash
smithery skill add whambammy/robotics-spatial-vision-mcp
```

### Claude Desktop / Cursor (`claude_desktop_config.json`)
```json
{
  "mcpServers": {
    "robotics-spatial-vision-mcp": {
      "command": "npx",
      "args": ["-y", "@whambammy/robotics-spatial-vision-mcp"],
      "env": {
        "PAYMENT_WALLET": "0x9793E7269b3301893318dEa8338576Ba612F39B3",
        "BASE_RPC_URL": "https://mainnet.base.org"
      }
    }
  }
}
```

---

## 🛠️ Included Tools

| Tool Name | Price (USDC) | Capability |
| :--- | :---: | :--- |
| `pointcloud_icp_rigid_registration` | $0.045 | Iterative Closest Point (ICP) algorithm aligning two unorganized 3D point clouds with robust Tukey loss weighting to filter noisy scan outliers. |
| `navmesh_recast_baking_validator` | $0.040 | Generates and audits Recast 3D navigation meshes for autonomous agents, checking maximum slope climbing, step-height clearance, and unreachable islands. |
| `robot_arm_inverse_kinematics_ccd` | $0.040 | Cyclic Coordinate Descent (CCD) and FABRIK solver computing 6-DOF robotic manipulator joint angles with strict joint-angle limit enforcement. |
| `lidar_ground_plane_ransac_filter` | $0.035 | RANSAC plane segmentation isolating ground terrain from foreground obstacles and vehicles in dense unorganized LIDAR point clouds. |
| `camera_intrinsics_projection_matrix` | $0.020 | Computes 4x4 perspective and orthographic camera projection matrices, mapping FOV, focal length, sensor dimensions, and near/far clipping planes. |


---

## 🔄 End-to-End Workflow

A robotics agent receives multi-sensor LIDAR and camera frames -> runs RANSAC ground plane segmentation -> aligns point clouds using ICP -> computes inverse kinematics to reach an obstacle-free target on a validated Recast navigation mesh.

---

## 💰 The x402 Base L2 Micropayment Protocol

When an agent invokes a tool without payment, the server responds with a deterministic `HTTP 402 Payment Required` challenge containing:
- Target tool price in USDC
- Base Native USDC Contract: `0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913`
- Recipient payout wallet address
- Single-use cryptographic nonce

Once broadcasted on Base L2, resubmitting with `paymentSignature` unlocks deterministic execution.

---

## 📄 License
MIT License. Created by [Whambammy](https://github.com/Whambammy).

TDQS

A3.7/5.0

Scored across 5 tools

Disambiguation5/5

Each tool targets a clearly distinct algorithmic task: point cloud registration, navmesh baking, inverse kinematics, ground plane filtering, and camera projection matrices. There is no overlap in purpose and an agent can select unambiguously from names alone.

Naming Consistency4/5

All names are consistently snake_case and descriptively compound (domain + algorithm/method), so style is uniform. The only weakness is that they are noun phrases rather than a predictable verb_noun pattern, and the internal ordering of qualifiers varies slightly across tools.

Tool Count5/5

Five tools is well within the ideal range and each maps to a distinct spatial-vision capability, so nothing feels padded or redundant. The set is tightly scoped to a specialized robotics domain.

Completeness4/5

The surface covers several core spatial perception tasks (registration, segmentation, kinematics, projection, navigation), giving a usable toolkit. However a broad 'Robotics & Spatial Vision' domain suggests missing operations like point-cloud I/O, coordinate transforms, forward kinematics, or path planning, which agents would need to work around.

Maintenance

ActivityMaintained
ResponsivenessNo issues