# CarryFeed [Health: Active]

**Category:** 🌐 Social Media  
**Repository:** https://github.com/christian-beep383/carryfeed-agent-tools  
**GitHub Stars:** 0  
**npm Downloads (last month):** 15  
**Views:** 1  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/carryfeed

## Description
Resolve public X/Twitter profiles, posts, and media through CarryFeed.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "carryfeed": {
    "command": "npx",
    "args": ["-y","@carryfeed/cli"]
  }
}
```

## Documentation

## What CarryFeed MCP server does

CarryFeed MCP server connects MCP-capable agents to the hosted CarryFeed API for reading public X/Twitter content. Its supported workflows cover public profiles, individual posts, articles, and attached media such as images, videos, and GIFs. Returned data can retain the original source URL together with author information, text, and media candidates, which helps an agent keep the result tied to its source.

The repository is a wrapper layer rather than the underlying resolver. It does not provide CarryFeed's internal implementation, worker code, direct X/Twitter client, authentication tokens, cookies, or scraping mechanisms.

## How it works

The MCP package communicates with CarryFeed's hosted service over a standard stdio MCP process. An MCP client starts the package with `npx`, then sends requests through the protocol. CarryFeed handles resolution of the supplied public X/Twitter identifier and returns the available metadata.

The same repository also publishes a JavaScript SDK and CLI. Their documented operations include resolving a post URL, resolving media from a post URL, and retrieving a profile by name. The MCP server is intended to provide these public-link workflows to agents through MCP rather than requiring each agent to integrate the API independently.

## Setup and configuration

Add the package to an MCP client's server configuration with a command of `npx` and the package argument `-y @carryfeed/mcp`:

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

The repository also provides `@carryfeed/sdk` for JavaScript applications and `@carryfeed/cli` for command-line or agent-script use. The CLI examples support compact output, profile status retrieval, and media resolution.

## Tools and capabilities

CarryFeed MCP server supports agent workflows for:

- Resolving public X/Twitter post links.
- Reading public profile information.
- Retrieving public article metadata.
- Finding image, video, and GIF-style media associated with public posts.
- Keeping source URLs and related author or text context with results.
- Using a hosted API instead of browser-driven X/Twitter access.

A separate download helper is documented for public `pbs.twimg.com` and `video.twimg.com` media URLs. The repository also includes OpenAPI notes, API health information, and an agent skill that directs agents toward the CLI for public X/Twitter source context.

## Limitations and notes

Only public X/Twitter links are supported. The service does not provide access to private or protected accounts, deleted content, restricted material, or pages that require login. Search and hashtag search are listed as future API work, so they should not be treated as current MCP capabilities.

Results depend on what the hosted CarryFeed API can resolve from the supplied public link. This package does not expose X/Twitter credentials or a direct scraping implementation. Use the server for public-source metadata workflows, not for authenticated account access or unrestricted platform search.

_Full upstream README: https://allmcps.com/mcp/carryfeed/readme_