bytesagain-mcp
# BytesAgain MCP Server
[](https://glama.ai/mcp/servers/bytesagain/mcp-server)
Search 60,000+ AI agent skills directly from any MCP-compatible agent.
## Overview
BytesAgain is a free MCP server for discovering AI agent skills and workflow use cases across ClawHub, LobeHub, Dify, GitHub-indexed skills, and the BytesAgain curated catalog.
This repository contains a real stdio MCP server wrapper. It exposes compact tools to agents and forwards read-only requests to the public BytesAgain API at `https://bytesagain.com/api/mcp`.
## MCP Tools
| Tool | When to use | Returns |
| --- | --- | --- |
| `search_skills` | Find AI skills for a concrete task, keyword, domain, or integration request. Supports English, Chinese, Japanese, Korean, German, French, Spanish, and Portuguese. | Ranked skill summaries with slug, name, description, category, tags, downloads, owner, and relevance fields. |
| `get_skill` | Fetch full details for a specific slug returned by `search_skills` or `popular_skills`. | Detailed skill metadata, install/source links, category, tags, owner, downloads, stars, and related fields when available. |
| `popular_skills` | Browse trending or high-download skills when the user has no specific task in mind. | Top skills ranked by downloads. |
| `search_use_cases` | Search workflow/use-case pages such as “write weekly reports”, “build dashboards”, or “automate ecommerce listings”. | Use-case pages and descriptions that map real-world tasks to relevant skills. |
## Install
### Run with npx
```bash
npx -y --package github:bytesagain/mcp-server bytesagain-mcp
```
### Run from source
```bash
git clone https://github.com/bytesagain/mcp-server.git
cd mcp-server
npm install
npm start
```
### Docker
```bash
docker build -t bytesagain-mcp .
docker run --rm -i bytesagain-mcp
```
## Claude Desktop configuration
```json
{
"mcpServers": {
"bytesagain": {
"command": "npx",
"args": ["-y", "--package", "github:bytesagain/mcp-server", "bytesagain-mcp"]
}
}
}
```
## Environment variables
| Variable | Default | Description |
| --- | --- | --- |
| `BYTESAGAIN_API_BASE` | `https://bytesagain.com/api/mcp` | Optional override for the public BytesAgain API endpoint. |
No API key is required. The server is read-only and does not write to BytesAgain, GitHub, Glama, or third-party services.
## Public endpoints
- Website: <https://bytesagain.com>
- MCP documentation: <https://bytesagain.com/mcp>
- Hosted MCP endpoint: <https://bytesagain.com/api/mcp>
- REST search example: <https://bytesagain.com/api/mcp?action=search&q=video%20editor>
## Development
```bash
npm install
npm test
```
The smoke test starts the MCP server over stdio and verifies that the tool list is exposed correctly.
## License
MIT
TDQS
Scored across 4 tools
Each tool serves a clearly distinct purpose: get_skill for fetching details by slug, popular_skills for browsing top skills, search_skills for keyword-based skill discovery, and search_use_cases for finding workflow-driven content. No overlap exists.
All tool names follow the verb_noun pattern using underscores (get_skill, popular_skills, search_skills, search_use_cases). While 'popular_skills' uses an adjective rather than a verb, it is consistent in style and easily understood.
With only 4 tools, the surface is lean yet sufficient for the domain of searching and retrieving AI skills and use-cases. Each tool contributes a necessary function without redundancy.
Core operations are covered: searching skills, getting skill details, listing popular skills, and searching use-cases. However, there is no tool to retrieve full details for a specific use-case, which is a minor gap given that get_skill only covers skills.