# prompts.chat-mcp [Health: Active]

**Category:** 🗣️ Conversational AI  
**Repository:** https://github.com/f/prompts.chat-mcp  
**GitHub Stars:** 40  
**npm Downloads (last month):** 1464  
**Views:** 12  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/prompts-chat-mcp

## Description
A MCP server for prompts.chat

## Tools
Capabilities this server exposes over MCP:

- **search_prompts** — Search for AI prompts by keyword. Returns matching prompts with title, description, content, author, category, and tags.
- **get_prompt** — Get a prompt by ID. If the prompt contains template variables (like ${variable} or ${variable:default}), you may need to provide values for them.

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

```json
"mcpServers": {
  "prompts-chat-mcp": {
    "command": "npx",
    "args": ["-y","@fkadev/prompts.chat-mcp"],
    "env": {
      "PROMPTS_API_KEY": "",
      "PROMPTS_QUERY": ""
    }
  }
}
```

**Requires environment variables:** `PROMPTS_API_KEY`, `PROMPTS_QUERY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What the prompts.chat-mcp MCP server does

The prompts.chat-mcp MCP server gives an AI coding assistant access to prompts.chat content through the Model Context Protocol. It can search the prompt library by keyword and filter results by type, category, or tag. Agents can retrieve an individual prompt by ID and provide values for template variables written with `${variable}` syntax.

The server also exposes public prompts through MCP's prompts capability. Clients can use `prompts/list` to browse available entries and `prompts/get` to retrieve a prompt with variable substitution. This makes the server useful when an assistant needs an existing prompt rather than generating one from scratch.

## How it works

You can connect to the hosted MCP endpoint at `https://prompts.chat/api/mcp`, or run the npm package locally over stdio. The hosted option avoids a local Node.js process and is shown as the recommended setup in the client examples. The local package is started with `npx -y @fkadev/prompts.chat-mcp` and requires Node.js version 18 or newer.

The prompts.chat-mcp MCP server communicates with the client through MCP. Search requests can include keyword and filtering information, while prompt retrieval uses an ID and can request values for declared template variables. Saving a prompt is available through the API-backed tool but requires an API key.

## Setup and configuration

For a remote connection, add `https://prompts.chat/api/mcp` to the MCP configuration for the client. Cursor uses `url`, Windsurf uses `serverUrl`, and VS Code configures the endpoint as an HTTP server. Claude Code supports the equivalent `claude mcp add --transport http` command.

For local use, configure the client to run `npx -y @fkadev/prompts.chat-mcp`. The repository provides local configuration examples for Cursor, Windsurf, VS Code, Claude Code, and Zed.

The local server accepts two environment variables:

- `PROMPTS_API_KEY` optionally authenticates requests and is required for `save_prompt`.
- `PROMPTS_QUERY` optionally filters the prompts exposed by the server. Its value can contain query parameters such as users, categories, and tags.

## Tools and capabilities

The prompts.chat-mcp MCP server provides these tools:

- `search_prompts`: Finds prompts by keyword, with filtering by type, category, or tag.
- `get_prompt`: Retrieves a prompt by ID and supports variable elicitation for templates.
- `save_prompt`: Saves a new prompt to the user's account; an API key is required.

In addition to tools, the server publishes public prompts through MCP's standard prompt operations. A client can browse with `prompts/list` and retrieve entries with `prompts/get`. Example requests include searching for code-review prompts or retrieving a prompt named “act as a linux terminal.”

## Limitations and notes

`PROMPTS_QUERY` is documented for the local server configuration; the material does not specify equivalent filtering options for the hosted endpoint. Saving prompts cannot be used without `PROMPTS_API_KEY`.

_Full upstream README: https://allmcps.com/mcp/prompts-chat-mcp/readme_

