# clarifyprompt

Turns a vague prompt into one shaped for a specific platform's syntax and parameters, using signals from your workspace.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 13
- Author: LumabyteCo
- License: Apache-2.0 · policy=link-only · description only; the file stays with the author
- Source: https://github.com/LumabyteCo/clarifyprompt-mcp
- Updated: 2026-10-06
- HTML: https://affpapa.org/en/nevibecode/mcp/clarifyprompt

## What it does

Send the raw request. The server reads workspace files such as `CLAUDE.md`, `AGENTS.md`, `.cursorrules` and `package.json`, works out the intent, and returns a prompt built for the platform you picked. The pack covers chat models, coding assistants, image and video generators; each has its own syntax and knobs.

There is `optimize_prompt` with a `platform` argument, a `critique_prompt` call that scores six dimensions, and an `auto_revise` loop that keeps fixing weak spots. Hosts supporting the `io.modelcontextprotocol/ui` extension get a `compose_prompt` panel with a before/after view and Accept / Revise buttons. Three environment variables are required: `LLM_API_URL`, `LLM_API_KEY`, `LLM_MODEL`. The server calls a model with your key, so it won't start without them. Node.js 18 or newer is needed.

Leave the platform out and the output stays neutral, without any vendor's idioms. Useful when one request goes to several tools.

## Requirements
- Needs: Node.js (commands run via npx)
- Environment variables: `LLM_API_URL`, `LLM_API_KEY`, `LLM_MODEL` — keys and tokens: ask the human for the values, never commit them or put them in shared files.

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add clarifyprompt -e LLM_API_URL=<LLM_API_URL> -e LLM_API_KEY=<LLM_API_KEY> -e LLM_MODEL=<LLM_MODEL> -- npx -y clarifyprompt-mcp
```
Replace <…> with your own values

### Codex

**Platform CLI**
```bash
codex mcp add clarifyprompt --env LLM_API_URL=<LLM_API_URL> --env LLM_API_KEY=<LLM_API_KEY> --env LLM_MODEL=<LLM_MODEL> -- npx -y clarifyprompt-mcp
```
Replace <…> with your own values

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "clarifyprompt": {
            "command": "npx",
            "args": [
                "-y",
                "clarifyprompt-mcp"
            ],
            "env": {
                "LLM_API_URL": "<LLM_API_URL>",
                "LLM_API_KEY": "<LLM_API_KEY>",
                "LLM_MODEL": "<LLM_MODEL>"
            }
        }
    }
}
```
Replace <…> with your own values

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "clarifyprompt": {
            "type": "stdio",
            "command": "npx",
            "args": [
                "-y",
                "clarifyprompt-mcp"
            ],
            "env": {
                "LLM_API_URL": "<LLM_API_URL>",
                "LLM_API_KEY": "<LLM_API_KEY>",
                "LLM_MODEL": "<LLM_MODEL>"
            }
        }
    }
}
```
Replace <…> with your own values

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add -e LLM_API_URL=<LLM_API_URL> -e LLM_API_KEY=<LLM_API_KEY> -e LLM_MODEL=<LLM_MODEL> clarifyprompt npx -y clarifyprompt-mcp
```
Replace <…> with your own values

## When to use it
- rewrite a vague prompt to match a target platform's syntax
- pull `CLAUDE.md` and `package.json` into the request text
- compare original and optimized prompts before sending
- score a prompt on six dimensions and revise it
- keep output neutral when no platform is chosen

## Example prompts
- Optimize this prompt for Midjourney: a ginger cat on a windowsill in morning light
- Call `optimize_prompt` with `platform: "gpt-image"` and keep the in-image text
- Run `critique_prompt` on my prompt and list what to fix

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 386989
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33918
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81115
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119160
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47392
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43872

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md