# PromptFrenzy

**Category:** 💻 Developer Tools  
**Repository:** https://github.com/Prompt-Frenzy/skills  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/promptfrenzy

## Description
Find a tested AI image or video prompt — toy, LEGO, plushie, portrait, poster — and render it.

## Claude Desktop Quick Installation
Heuristic fallback — verify the package name and runner against the repository README before running it. Uses `npx` (confidence: low):

```json
"mcpServers": {
  "promptfrenzy": {
    "command": "npx",
    "args": ["-y","promptfrenzy"]
  }
}
```

## Documentation & README

# PromptFrenzy skills

Agent skills for [PromptFrenzy](https://www.promptfrenzy.com) — a library of AI
image and video prompts that have each been run thousands of times, plus the
generator that runs them.

## `promptfrenzy`

Use when someone wants a picture or clip made from a photo or an idea — a LEGO
or toy version, a plushie, a portrait, a poster, a product shot, a character
sheet, or a short animated clip — and wants a prompt that already works rather
than one written from scratch.

The skill teaches an agent to search the library, read what a prompt needs
before spending anything, and either hand the prompt to the user or render it
for them through the [MCP server](https://www.promptfrenzy.com/mcp).

### Install

```bash
npx skills add Prompt-Frenzy/skills
```

Or copy `promptfrenzy/` into your client's skills directory.

### Connect the generator

Searching and reading prompts need no account. To render, add the MCP server:

```bash
claude mcp add --transport http promptfrenzy \
  https://www.promptfrenzy.com/api/mcp \
  --header "Authorization: Bearer pf_your_key_here"
```

Keys are created at [Settings → Connect an AI agent](https://www.promptfrenzy.com/settings).
Full docs: <https://www.promptfrenzy.com/mcp>.

## Source of truth

These files are maintained in the PromptFrenzy product repo alongside the
server they document, and mirrored here. Please open issues here; fixes land
upstream first so the skill and the API can't drift into two different truths.

MIT licensed.