About Moonscroll
Moonscroll is a hosted remote MCP server for researching and analyzing short-form video content across Instagram Reels, TikTok, and YouTube Shorts. It supports use cases such as trend research, creator research, competitive briefs, and content ideation.
Key features
- 01Insights for Instagram Reels
- 02Insights for TikTok
- 03Insights for YouTube Shorts
- 04Remote Streamable HTTP endpoint
Use cases
- 01Spot trends in short-form videos
- 02Research creators and competitors
- 03Develop content ideas from public performance signals
MCP support
MCP clients connect to the hosted server at https://api.moonscroll.ai/mcp over Streamable HTTP. The tools and their parameters are defined by the live server.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://api.moonscroll.ai/mcpclaude mcp add --transport http thefedoration-moonscroll https://api.moonscroll.ai/mcp{"mcpServers":{"thefedoration-moonscroll":{"type":"http","url":"https://api.moonscroll.ai/mcp"}}}Open in Cursor{"servers":{"thefedoration-moonscroll":{"type":"http","url":"https://api.moonscroll.ai/mcp"}}}Open in VS Code{"mcpServers":{"thefedoration-moonscroll":{"type":"http","url":"https://api.moonscroll.ai/mcp"}}}{"mcpServers":{"thefedoration-moonscroll":{"serverUrl":"https://api.moonscroll.ai/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
13| Parameter | Type | Required | Description |
|---|---|---|---|
| limit | integer | No |
Details
From the project READMEMoonscroll MCP
Short Form Video Insights for Instagram, TikTok, Shorts
This repository packages the public, remote Moonscroll Model Context Protocol (MCP) server for easy install in Claude, Cursor, VS Code, and related MCP clients. The backend is hosted by Moonscroll — this repo does not include or open-source the server implementation.
| MCP endpoint | https://api.moonscroll.ai/mcp |
| Website | https://moonscroll.ai/ |
| API docs | https://api.moonscroll.ai/docs/ |
| Privacy | https://moonscroll.ai/privacy-policy/ |
| Transport | Streamable HTTP (remote only) |
TODO: Replace every
YOUR_GITHUB_USERNAMEplaceholder in this repo (server.json,.cursor-plugin/plugin.json,glama.json, and any README links) with your real GitHub username or org before publishing to the Official MCP Registry, Glama, or GitHub.
What it does
Moonscroll exposes short-form video insight tools over MCP so agents can research and analyze content on:
- Instagram Reels
- TikTok
- YouTube Shorts
Typical use cases: trend spotting, creator research, competitive briefs, content ideation, and summarizing public short-form performance signals. Exact tools and parameters are defined by the live remote server — see the API docs.
Remote-only packaging
- There is no local Node/Python server to run from this repo.
- There are no API keys checked in. If the hosted MCP requires auth (OAuth or a bearer token), configure that in your client — never commit secrets.
- Clients connect directly to
https://api.moonscroll.ai/mcpover Streamable HTTP.
Install
Cursor (recommended: plugin / Marketplace)
This repo is structured as a Cursor plugin (.cursor-plugin/plugin.json + root mcp.json).
- Install from the Cursor Marketplace when listed, or clone this repo and add it as a local/team plugin.
- Ensure the Moonscroll MCP server appears under MCP settings and points at
https://api.moonscroll.ai/mcp. - Complete any sign-in / auth flow the client prompts for.
Manual Cursor config (~/.cursor/mcp.json or project .cursor/mcp.json):
{
"mcpServers": {
"moonscroll": {
"url": "https://api.moonscroll.ai/mcp"
}
}
}
Optional headers (only if Moonscroll documents a static token — do not invent keys):
{
"mcpServers": {
"moonscroll": {
"url": "https://api.moonscroll.ai/mcp",
"headers": {
"Authorization": "Bearer ${env:MOONSCROLL_TOKEN}"
}
}
}
}
Claude Code
claude mcp add --transport http moonscroll https://api.moonscroll.ai/mcp
Or project .mcp.json:
{
"mcpServers": {
"moonscroll": {
"type": "http",
"url": "https://api.moonscroll.ai/mcp"
}
}
}
Claude Desktop (custom connector)
Add a remote MCP connector with URL https://api.moonscroll.ai/mcp in Claude Desktop settings (Custom Connectors / MCP). Prefer OAuth when offered by the host. For header-only auth, follow Anthropic’s current remote MCP guidance.
VS Code
Create or edit .vscode/mcp.json:
{
"servers": {
"moonscroll": {
"type": "http",
"url": "https://api.moonscroll.ai/mcp"
}
}
}
Official MCP Registry (server.json)
server.json declares this as a remote registry entry with Streamable HTTP:
"remotes": [
{
"type": "streamable-http",
"url": "https://api.moonscroll.ai/mcp"
}
]
Namespace placeholder: io.github.YOUR_GITHUB_USERNAME/moonscroll — replace YOUR_GITHUB_USERNAME before publishing.
Agent skill
skills/short-form-video-insights/SKILL.md is a starter skill for agents using Moonscroll for Instagram / TikTok / Shorts research. Keep tool calls aligned with the live server’s tool list from the MCP session.
Directory layout
moonscroll-mcp/
├── .cursor-plugin/plugin.json # Cursor Marketplace manifest
├── assets/logo.svg # Brand mark (from moonscroll.ai/favicon.svg)
├── skills/short-form-video-insights/SKILL.md
├── glama.json # Glama ownership claim (maintainers placeholder)
├── mcp.json # Cursor plugin MCP (remote URL)
├── server.json # Official MCP Registry remote server
├── smithery.yaml # Minimal Smithery remote HTTP config
├── LICENSE # MIT
├── README.md
└── .gitignore
License
MIT — see LICENSE. The hosted Moonscroll API and product remain subject to Moonscroll’s terms and privacy policy.
Links
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.