banana-prompts-mcp
OfficialREADME.md
# Banana Prompts MCP Server
This MCP (Model Context Protocol) server allows you to search for high-quality AI art prompts directly from [Banana Prompts](https://bananaprompts.fun).
Explore a curated collection of Nano Banana prompts to inspire your next creation.
## Features
- **Search by Tags**: Find prompts by specific categories or styles (e.g., "Cyberpunk", "Portrait", "Oil Painting").
- **Search by Query**: Perform a general keyword search across prompt titles and content.
- **Natural Integration**: Each result includes a direct link to the full prompt and high-resolution images on the [Banana Prompts website](https://bananaprompts.fun).
## Installation
To use this server with Claude Desktop or any other MCP-compatible client, follow these steps:
### 1. Requirements
- [Node.js](https://nodejs.org/) (v18 or higher recommended)
- An MCP-compatible client (like Claude Desktop)
### 2. Setup
Clone the repository or download the files:
```bash
git clone https://github.com/Banana-Prompts/banana-prompts-mcp.git
cd banana-prompts-mcp
npm install
npm run build
```
### 3. Configuration
#### Claude Desktop
Add the following to your `claude_desktop_config.json`:
```json
{
"mcpServers": {
"banana-prompts": {
"command": "node",
"args": ["/path/to/banana-prompts-mcp/dist/index.js"]
}
}
}
```
Replace `/path/to/banana-prompts-mcp/dist/index.js` with the actual absolute path to the compiled `dist/index.js` file.
## Usage
Once configured, you can ask Claude to:
- "Find some cyberpunk prompts on Banana Prompts"
- "Search Banana Prompts for sunset landscape prompts"
- "Show me prompts tagged with 'watercolor' from bananaprompts.fun"
## Available Tools
- `search_prompts_by_tags`: Input an array of tags to filter prompts.
- `search_prompts_by_query`: Input a search string to find matching prompts.
---
Built by the team at [Banana Prompts (bananaprompts.fun)](https://bananaprompts.fun).
TDQS
B3.4/5.0
Scored across 2 tools
Disambiguation5/5
The two tools have clearly distinct purposes: one searches by a general query, the other by tags. There is no overlap in functionality, making them easy to distinguish for an agent.
Naming Consistency5/5
Both tools follow the consistent verb_noun pattern 'search_prompts_by_<criterion>', making the naming predictable and intuitive.
Tool Count2/5
Only two tools is very few for a server. Even though it focuses on search, the minimal count suggests an underdeveloped surface that may not meet diverse agent needs.
Completeness2/5
Only search operations are provided; there are no tools for creating, updating, or deleting prompts. This leaves obvious gaps for any workflow beyond simple lookup.
Maintenance
ActivityInactive
ResponsivenessNo issues