glim.sh - live data for any agent (hosted, Streamable HTTP)

Name: glim.sh - live data for any agent (hosted, Streamable HTTP)
Author: glim-sh

@glim-sh

代码仓库主页文档

关于 glim.sh - live data for any agent (hosted, Streamable HTTP)

Hosted MCP server with data tools for AI agents. Connect at https://glim.sh/mcp.

基本信息

分类

数据与分析

许可证

MIT

发布者

glim-sh

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "glim": {
      "type": "http",
      "url": "https://glim.sh/mcp"
    }
  }
}

工具

$0.005

$0.01

$0.015

$0.01

$0.002

$0.005

$0.01

概览

What is glim.sh - live data for any agent (hosted, Streamable HTTP)?

glim.sh - live data for any agent (hosted, Streamable HTTP) is a hosted MCP server endpoint that gives AI agents live data from Twitter/X, Reddit, the open web, GitHub, Amazon, and YouTube without requiring API keys or a scraping stack. It charges per call and supports both account‑based billing and crypto wallet auto‑pay.

How to use glim.sh - live data for any agent (hosted, Streamable HTTP)?

Add the endpoint https://glim.sh/mcp to any MCP‑compatible client (Claude, Codex, OpenCode, etc.) through a one‑click link, a CLI command, or a JSON config. On first use the client runs a browser OAuth sign‑in. Manage your balance at glim.sh/app. Alternatively, use a crypto wallet path: calls auto‑pay USDC (Base or Solana) via x402 challenges; MPP on Tempo is also accepted on the same endpoint. Tools can be restricted per session by appending ?tools= to the URL.

Key features of glim.sh - live data for any agent (hosted, Streamable HTTP)

One endpoint provides tools for live data from six platforms.
No API keys or scraping infrastructure needed.
Pay‑per‑call pricing with per‑tool costs shown in the README.
Supports both account‑based (OAuth sign‑in) and crypto wallet (x402) payment.
Tool list can be narrowed per session via ?tools= parameter.
Every MCP tool is also available as a plain REST API at https://glim.sh/api/v1/.

Use cases of glim.sh - live data for any agent (hosted, Streamable HTTP)

Search and retrieve tweets with advanced operators (e.g., for social media monitoring).
Fetch Reddit posts, comments, subreddit feeds, or user activity.
Perform semantic web searches and clean page extraction (including SSR, SPA shells, PDFs).
Search GitHub repos, conversations, code, and retrieve repo metadata and PRs.
Search Amazon listings with prices, ratings, and ASINs, and get product details.
Get YouTube subtitles or transcripts by video ID.

FAQ from glim.sh - live data for any agent (hosted, Streamable HTTP)

How do I connect my agent to glim.sh - live data for any agent (hosted, Streamable HTTP)?

Add the endpoint URL https://glim.sh/mcp to your MCP client’s configuration. For common clients (Claude, Codex, OpenCode, etc.), the README provides one‑click links, CLI commands, or a JSON config block.

What data sources and tools are available?

The server exposes tools for Twitter/X (search, get), Reddit (search, get), web (search, fetch), GitHub (search, get), Amazon (search, get), and YouTube (get transcripts). Each tool has a fixed price per call; see the tool table in the README.

How does billing work?

Billing is pay‑per‑call. You can sign in with a browser OAuth and manage your balance at glim.sh/app. Alternatively, use a crypto wallet path: calls auto‑pay USDC (Base or Solana) via x402 challenges, and MPP on Tempo is also accepted on the same endpoint.

Can I restrict which tools the agent can use?

Yes. Append ?tools= followed by a comma‑separated list of tool names to the endpoint URL. For example: https://glim.sh/mcp?tools=glim_twitter_search,glim_web_fetch.

Is there a REST API in addition to the MCP endpoint?

Yes. Every MCP tool is mirrored as a REST endpoint under https://glim.sh/api/v1/, g

数据与分析分类下的更多 MCP 服务器

HubSpot MCP Server

baryhuang

A Model Context Protocol (MCP) server that enables AI assistants to interact with HubSpot CRM data, providing built-in vector storage and caching mechanisms help overcome HubSpot API limitations while improving response times.

128