Perplexity AI MCP Server
# Perplexity AI MCP Server
This repository contains the source code for a Model Context Protocol (MCP) server that provides access to the Perplexity AI API. This server allows users to interact with Perplexity AI through various tools, including chatting, searching, and retrieving documentation.
## Purpose
This server simplifies the integration of Perplexity AI into MCP-based systems. It provides a convenient and standardized way to access Perplexity AI's capabilities.
## Setup
1. **Install Node.js and npm:** Ensure you have Node.js and npm installed on your system.
2. **Clone the repository:** Clone this repository to your local machine.
3. **Install dependencies:** Navigate to the project directory and run `npm install`.
4. **Configure API Key:** Set the `PERPLEXITY_API_KEY` environment variable to your Perplexity API key.
5. **Run the server:** Run `npm start` to start the server.
## Usage
The server exposes several tools that can be accessed through the MCP system. Refer to the MCP documentation for details on how to use these tools.
## Technologies Used
* TypeScript
* @modelcontextprotocol/sdk
* axios
## Known Issues
* The Perplexity API may be unreliable. Error handling is included to gracefully handle API failures.
## Contributing
Contributions are welcome! Please open an issue or submit a pull request.
TDQS
Scored across 5 tools
Most tools have distinct purposes: chat_perplexity handles conversational AI, check_deprecated_code analyzes code deprecation, find_apis discovers APIs, get_documentation retrieves docs, and search performs general queries. However, get_documentation and search could overlap slightly when users seek technical information, but descriptions clarify their focus.
Naming is mixed: chat_perplexity uses a verb_noun format, while check_deprecated_code, find_apis, get_documentation, and search use verb-based phrases without a consistent pattern. All names are snake_case, providing some readability, but the verb styles vary (e.g., 'chat' vs. 'check' vs. 'find'), lacking a unified convention.
With 5 tools, the count is reasonable for a server focused on AI assistance and information retrieval. It covers key areas like conversation, code analysis, API discovery, documentation, and general search, though it might feel slightly thin for broader AI tasks, but each tool earns its place.
The server targets AI-driven information and code assistance, with tools for chat, code checks, API finding, documentation, and search. Notable gaps include lack of update/delete operations for chats or saved searches, and no tool for summarizing or analyzing search results, which could limit agent workflows in this domain.