# optics-mcp [Health: Active]

**Category:** 💰 Finance & Fintech  
**Repository:** https://github.com/sunex-ai/optics-mcp  
**GitHub Stars:** 0  
**npm Downloads (last month):** 204  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/optics-mcp

## Description
Search Sunex lenses and imager sensors, match compatible optics, get pricing. No API key needed.

## Claude Desktop Quick Installation
Heuristic fallback — verify the package name and runner against the repository README before running it. Uses `npx` (confidence: low):

```json
"mcpServers": {
  "optics-mcp": {
    "command": "npx",
    "args": ["-y","optics-mcp"]
  }
}
```

## Documentation

## What the optics-mcp MCP server does

The optics-mcp MCP server exposes Sunex's lens and imager catalog to MCP-compatible AI clients. It supports product discovery, sensor specification lookup, compatibility checks, and basic optical geometry calculations. Results can include lens field of view, angular resolution, sensor dimensions, sample pricing, and links for requesting a quote.

The server is read-only. It does not provide catalog mutations, ordering, or account-specific workflows. Reach for it when an agent needs to turn requirements such as sensor part number, pixel pitch, resolution, focal coverage, or maximum f-number into candidate Sunex optics.

## How it works

Clients connect to the hosted endpoint at `https://mcp.sunex-ai.com/mcp` using Streamable HTTP, the transport specified by the MCP standard version 2025-03-26. An older Server-Sent Events endpoint remains available at `/sse`. The service runs as a thin Cloudflare Workers proxy over the Sunex production API at `optics-online.com/api/v1`.

The optics-mcp MCP server also publishes a discovery manifest at `https://mcp.sunex-ai.com/.well-known/mcp.json`. A landing page at the service root contains connection instructions. Claude can connect through a custom connector, while Cursor and other clients that support Streamable HTTP can use the endpoint in their MCP configuration. ChatGPT requires an MCP-to-OpenAPI bridge and a custom GPT Action.

## Tools and capabilities

Five tools are available:

- `recommend_lens_for_imager` accepts an imager part number and returns compatible lenses with field-of-view and angular-resolution information. Optional requirements can include a maximum f-number.
- `search_imagers` searches sensors by part number, manufacturer, or resolution class.
- `get_imager_detail` returns full sensor specifications and calculates width, height, and diagonal in millimeters.
- `find_compatible_lenses` uses pixel count and pixel pitch to find lenses whose image circle covers the sensor. Queries can also specify field-of-view requirements.
- `search_products` searches the broader catalog by part number or keyword and can return sample prices and RFQ links.

Example tasks include finding a wide-angle lens for the Sony IMX577, locating fisheye lenses below a target price, calculating the diagonal of an IMX477, or matching lenses to a 1920×1080 sensor with 3-micron pixels and a specified horizontal field of view.

## Setup and configuration

The hosted service needs no API key. Add `https://mcp.sunex-ai.com/mcp` as a custom MCP connection using Streamable HTTP. Clients that only support the older transport can use `https://mcp.sunex-ai.com/sse`.

The project is also self-hostable. The documented process is to clone the repository, run `npm install`, authenticate Wrangler with `npx wrangler login`, and deploy with `npx wrangler deploy`. The README does not specify additional required environment variables or authentication settings for the deployed server. The repository is licensed under MIT.

_Full upstream README: https://allmcps.com/mcp/optics-mcp/readme_