viralnote-mcp
Gives Claude, Cursor, and any MCP client 15 tools to create, schedule, and publish social posts to TikTok, Instagram Reels, YouTube Shorts, X, Threads, and LinkedIn. Also handles media import by URL or base64, webhooks, and per-platform analytics from the ViralNote API.
README
@viralnote/mcp-server
A Model Context Protocol server for the ViralNote social media API.
Plug it into Claude Desktop, Claude Code, Cursor, or any other MCP-aware host and your agent can schedule posts, manage media, and read analytics across X, Instagram, Facebook, TikTok, LinkedIn, YouTube, Pinterest, Bluesky, Threads, and Reddit — as native MCP tool calls. No glue code.
Install
Claude Desktop / Claude Code / Cursor
Add to your MCP config (~/Library/Application Support/Claude/claude_desktop_config.json on macOS, similar on other platforms):
{
"mcpServers": {
"viralnote": {
"command": "npx",
"args": ["-y", "@viralnote/mcp-server"],
"env": {
"VIRALNOTE_API_KEY": "vnd_..."
}
}
}
}
Restart your MCP host. The ViralNote tools will be available immediately.
This repo includes a root .mcp.json (Open Plugins standard) so tools like Cursor Directory can auto-detect the MCP server from the GitHub URL.
Local install
npm install -g @viralnote/mcp-server
Then reference viralnote-mcp directly in your MCP config:
{
"mcpServers": {
"viralnote": {
"command": "viralnote-mcp",
"env": { "VIRALNOTE_API_KEY": "vnd_..." }
}
}
}
Configuration
| Env var | Required | Default | Notes |
|---|---|---|---|
VIRALNOTE_API_KEY |
yes | — | Generate at viralnote.app/developers/auth. Grant posts:read, posts:write, plus webhooks:* if your agent should manage webhooks. |
VIRALNOTE_API_BASE |
no | https://viralnote.app/api/v1 |
Override for staging/self-hosted instances. |
Tools exposed
| Tool | Purpose |
|---|---|
list_posts |
List posts (filter by status/platform, paginated) |
get_post |
Read one post including per-platform publish results |
create_post |
Create a draft (is_draft: true) or scheduled post |
update_post |
Update a draft or scheduled post |
delete_post |
Delete (cancels if scheduled) |
publish_post |
Publish a draft now |
list_media |
List media library items |
import_media |
Import by URL (200MB) or base64 data (3MB) |
delete_media |
Delete a media item |
list_social_accounts |
List connected social accounts |
list_analytics |
Published posts with per-platform metrics |
list_post_results |
Per-platform delivery results (success/error) |
list_webhooks |
List webhook subscriptions |
create_webhook |
Subscribe to events |
delete_webhook |
Unsubscribe |
For most users, the HTTP MCP server at https://viralnote.app/api/mcp/mcp is simpler than installing this stdio package — see https://viralnote.app/developers/mcp for the HTTP config snippet. Use this stdio package when your MCP client doesn't support HTTP transport.
The underlying REST endpoints and request/response shapes are documented at viralnote.app/developers/docs.
Example agent prompts
"Show me my last 5 scheduled posts." Tool:
list_postswith{ status: "scheduled", limit: 5 }.
"Schedule this caption to Instagram for tomorrow at 9am, attaching the photo I uploaded yesterday." Tools:
list_media→ find item →create_postwith{ platforms: ["instagram"], caption, libraryItemId, scheduledFor, status: "scheduled" }.
"Pull this Dropbox link into my library, then publish it to X immediately." Tools:
import_media→create_post(draft) →publish_post.
Development
git clone https://github.com/viralnote/mcp-server
cd mcp-server
npm install
npm run build
VIRALNOTE_API_KEY=vnd_... npm start
For local iteration without rebuilding:
VIRALNOTE_API_KEY=vnd_... npm run dev
License
MIT — see LICENSE. Pull requests welcome.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。