Snap_sense

Name: Snap_sense
Author: Abhi5h3k

@Abhi5h3k

リポジトリ

Snap_sense について

Demo MCP tool to capture URL screenshot.

#snap-sense #mcp #screenshot-capture

基本情報

カテゴリ

その他

ランタイム

python

トランスポート

stdio

公開者

Abhi5h3k

投稿者

Abhishek Bhardwaj

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "snap_sense": {
      "command": "uv",
      "args": [
        "--directory",
        "/ABSOLUTE/PATH/TO/PARENT/FOLDER/weather",
        "run",
        "server.py"
      ],
      "env": {
        "ABSTRACT_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Snap_sense?

Snap_sense is a lightweight Model Context Protocol (MCP) server that enables your LLM to capture screenshots of any specified URL and return only the access URL for the captured image. It is built with Python and the MCP SDK, designed for integrating visual capture capabilities into AI applications like Claude Desktop or automation workflows.

How to use Snap_sense?

Install via Smithery (npx -y @smithery/cli install @Abhi5h3k/MCP-URL2SNAP --client claude) or manually: clone the repository, set up a Python virtual environment (Python 3.11+, UV 0.6.9+), install dependencies with uv install, create a .env file with your ABSTRACT_API_KEY, and run uv run server.py. Register the server in Claude Desktop's claude_desktop_config.json using the uv command and restart Claude. Then prompt with requests like "can you show me the screenshot of https://example.com".

Key features of Snap_sense

Lightweight MCP server for capturing webpage screenshots
Returns only the access URL of the captured image
Seamless integration with MCP-compatible LLMs
Easy setup with Python and the MCP SDK
Configurable via environment variable (AbstractAPI key)

Use cases of Snap_sense

Enabling AI assistants like Claude Desktop to capture and share webpage snapshots on demand
Automating screenshot generation for website monitoring or reporting workflows
Integrating visual capture into custom MCP-based automation pipelines

FAQ from Snap_sense

What dependencies are required to run Snap_sense?

The server requires Python 3.11.0 or higher and UV 0.6.9 or higher. All other dependencies are installed automatically via UV.

How is the screenshot captured and stored?

Snap_sense uses the AbstractAPI Screenshot API (accessed via your ABSTRACT_API_KEY) to capture the screenshot. The server returns only the access URL for the hosted image; the image itself is stored by AbstractAPI.

What transport and authentication does Snap_sense use?

The server communicates via the MCP protocol over standard I/O (stdio). Authentication to the screenshot service is provided through the ABSTRACT_API_KEY environment variable.

Can Snap_sense be used with any MCP-compatible client?

Yes, the README demonstrates integration with Claude Desktop via the claude_desktop_config.json file, and the server is designed to work with any MCP-compatible LLM or client that supports the protocol.