Presearch MCP Server

<div align="center"> <h1>🔍 Presearch MCP Server</h1> <p> <a href="https://smithery.ai/server/@NosytLabs/presearch-search-api-mcp"><img src="https://smithery.ai/badge/@NosytLabs/presearch-search-api-mcp" alt="Smithery Badge" /></a> <a href="https://presearch.io/"><img src="https://img.shields.io/badge/Presearch-Decentralized%20Search-blue?logo=presearch" alt="Presearch" /></a> <a href="https://modelcontextprotocol.io/"><img src="https://img.shields.io/badge/MCP-Model%20Context%20Protocol-green" alt="MCP" /></a> <a href="https://opensource.org/licenses/MIT"><img src="https://img.shields.io/badge/License-MIT-yellow.svg" alt="License" /></a> <a href="https://badge.fury.io/js/presearch-mcp-server"><img src="https://badge.fury.io/js/presearch-mcp-server.svg" alt="npm version" /></a> <a href="https://nodejs.org/"><img src="https://img.shields.io/badge/node-%3E%3D20.0.0-brightgreen" alt="Node.js" /></a> </p> <a href="https://presearch.com/signup?rid=4779685" target="_blank"> <img src="https://assets.presearch.com/referral/ban-4.jpg" title="Presearch" alt="Presearch Banner" /> </a> <h3>🔐 Privacy-First AI Search Integration</h3> <p><strong>Empower your AI with decentralized, uncensored web search capabilities through the Model Context Protocol (MCP).</strong></p> <a href="https://smithery.ai/server/@NosytLabs/presearch-search-api-mcp" target="_blank"> <img src="https://smithery.ai/button/deploy-on-smithery" alt="Deploy on Smithery" height="40" /> </a> </div> --- ## 🌟 Overview The **Presearch MCP Server** is a professional-grade integration bridge that connects AI assistants (like Claude, Cursor, and Trae) to the **Presearch decentralized search engine**. Unlike traditional search APIs that track user behavior, Presearch offers a decentralized, privacy-centric alternative. This server enables your AI to: <div align="center"> | 🛡️ **Search Anonymously** | 🔍 **Scrape Intelligently** | 🧠 **Research Deeply** | 📊 **Monitor Nodes** | |:--:|:--:|:--:|:--:| | No IP tracking or search history logging | Extract clean content from modern, dynamic websites | Perform multi-step investigations autonomously | Track the status and earnings of Presearch nodes | </div> --- ## 🔥 Real-World Examples See what's possible when you give your AI access to Presearch. ### 1. 🧠 Deep Research Mode **User Prompt:** *"Research the effects of climate change on coastal real estate markets."* **Tool Used:** `presearch_deep_research` ```json { "query": "climate change effects on coastal real estate", "breadth": 5, "depth": 2, "research_focus": "market" } ``` **Result:** The agent autonomously performs a multi-step investigation: 1. **Initial Search:** Queries Presearch for broad market trends. 2. **Analysis:** Identifies key sub-topics (insurance rates, flood risk zones, property devaluation). 3. **Deep Dive:** Executes targeted sub-searches for each topic. 4. **Synthesis:** Returns a comprehensive report citing 14+ distinct sources, including academic papers and market reports, with no tracking of the search intent. ### 2. ⚡ Search & Scrape (Fast Context) **User Prompt:** *"Get me the latest specs for the iPhone 16 Pro and summarize the camera upgrades."* **Tool Used:** `presearch_search_and_scrape` ```json { "query": "iPhone 16 Pro camera specs official", "limit": 3 } ``` **Result:** 1. **Search:** Finds the top 3 most relevant pages (Apple Official, TechCrunch, The Verge). 2. **Scrape:** Immediately fetches the full HTML content of all 3 pages in parallel using a headless browser. 3. **Output:** Returns 35kb of clean text content. 4. **AI Action:** The AI reads the raw specs and generates a perfect summary of the 48MP Ultra Wide camera and 5x Telephoto lens features. ### 3. 🛡️ Privacy-First Market Analysis **User Prompt:** *"Find competitors to our new SaaS product 'StealthMode' without alerting them via analytics."* **Tool Used:** `presearch_ai_search` ```json { "query": "StealthMode SaaS competitors", "safesearch": "strict", "freshness": "month" } ``` **Result:** * **Anonymity:** The searches are routed through Presearch's decentralized node network. The competitor websites see traffic from generic Presearch nodes, not your corporate IP address. * **Outcome:** A list of 10 direct competitors launched in the last month, gathered without leaving a digital footprint. --- ## 🛡️ What is Presearch? Presearch is a **decentralized search engine** built on blockchain technology that rewards community members with Presearch tokens (PRE) for their usage, contribution to, and promotion of the platform. ### Why Presearch Matters for AI <div align="center"> | 🚫 **Uncensored Access** | 🔒 **Privacy Protection** | 🌐 **Community Infrastructure** | |:--:|:--:|:--:| | Results are not filtered by central authorities | Your AI's queries are not profiled by ad-tech giants | Search index powered by independent community nodes | | Complete view of the web | Proprietary data remains private | Resilient, distributed control | </div> --- ## 💡 Key Features ### 🛡️ Privacy & Security - **Decentralized Infrastructure**: Leverages Presearch's distributed node network - **Bearer Token Auth**: Secure, standard authentication for API access - **No Data Persistence**: The server is stateless; no user queries are stored on disk ### 🔧 Robust Tooling - **Deep Research Mode**: Recursive search and analysis capabilities - **Smart Scraping**: Headless browser integration to scrape dynamic JS-heavy websites - **Flexible Input Handling**: Tools accept JSON strings and loose types for maximum compatibility with LLMs - **Multi-Format Export**: Export results to JSON, CSV, Markdown, HTML, or PDF ### 🚀 Enterprise Ready - **Intelligent Caching**: Configurable TTL and memory limits - **Rate Limiting & Retries**: Robust error handling with exponential backoff - **Health Monitoring**: Real-time status checks for API connectivity --- ## 🛠️ Available Tools <div align="center"> | Tool Name | Description | Key Parameters | |:----------|:------------|:---------------| | **`presearch_ai_search`** | Standard web search optimized for AI | `query`, `count`, `safesearch`, `freshness`, `content_categories` | | **`presearch_deep_research`** | Autonomous multi-step research agent | `query`, `depth`, `breadth`, `focus`, `location` | | **`presearch_search_and_scrape`** | Search and immediately scrape top results | `query`, `scrape_count`, `include_text`, `location` | | **`scrape_url_content`** | Scrape content from specific URLs | `urls`, `include_text`, `timeout_ms` | | **`analyze_content`** | Analyze content quality and relevance | `content`, `include_quality_assessment`, `custom_keywords` | | **`export_search_results`** | Export search results to files | `count`, `format` (json/csv/md/html/pdf), `file_output` | | **`presearch_site_export`** | Advanced export with scraping and analysis | `query`, `format`, `file_output`, `include_analysis`, `scrape_content` | | **`presearch_node_status`** | Monitor Presearch node health | `node_api_key`, `stats`, `connected`, `include_inactive` | | **`cache_stats`** | View internal cache metrics | (None) | | **`cache_clear`** | Clear the internal cache | (None) | | **`presearch_health_check`** | Verify API connectivity | (None) | </div> > **Note**: All tools support robust input parsing. Parameters can be passed as native types (numbers, booleans, arrays) or as strings/JSON strings (e.g., `"true"`, `"10"`, `"['url1', 'url2']"`). --- ## 📝 Available Prompts The server provides built-in prompts to help you get started with common tasks: <div align="center"> | Prompt Name | Purpose | |:------------|:--------| | **`presearch-deep-dive`** | Conduct deep research on a specific topic | | **`presearch-news`** | Find the latest news about a topic from the last 24 hours | | **`presearch-fact-check`** | Verify a claim or statement with evidence | | **`presearch-market-analysis`** | Analyze a market sector or product category | | **`presearch-node-monitor`** | Check the status and earnings of your Presearch nodes | | **`presearch-product-review`** | Research reviews and sentiment for a product | | **`presearch-academic`** | Conduct academic research prioritizing .edu and journals | | **`presearch-tutorial`** | Learn how to use a specific tool effectively | </div> --- ## 📚 Resources The server exposes the following resources for configuration and debugging: <div align="center"> | Resource URI | Description | |:-------------|:------------| | **`presearch://config`** | View current server configuration (secrets masked) | | **`presearch://rate-limits`** | Check current API rate limit status | | **`presearch://supported-countries`** | List of supported ISO 3166-1 alpha-2 country codes | | **`presearch://supported-languages`** | List of supported BCP 47 language codes | </div> --- ## ⚙️ Configuration The server can be configured via environment variables or MCP settings. ### Environment Variables <div align="center"> | Variable | Description | Default | |:---------|:------------|:--------| | `PRESEARCH_API_KEY` | Your Presearch API Key (Required) | - | | `PRESEARCH_BASE_URL` | API Endpoint URL | `https://na-us-1.presearch.com` | | `PRESEARCH_TIMEOUT` | Request timeout in ms | `10000` | | `LOG_LEVEL` | Logging verbosity (`info`, `debug`, `error`) | `info` | </div> ### JSON Configuration Schema When using Smithery or an MCP client, the configuration object supports: ```json { "apiKey": "YOUR_KEY", "rateLimit": { "maxRequests": 100, "windowMs": 60000 }, "cache": { "enabled": true, "ttl": 300 }, "search": { "defaultSafeSearch": "moderate", "defaultLanguage": "en-US" } } ``` --- ## 🚀 Quick Start ### 1. Get an API Key Sign up at [Presearch.io](https://presearch.io) to obtain your API key. ### 2. Run with npx ```bash npx presearch-mcp-server ``` ### 3. Deploy via Smithery Use the button above or run: ```bash npx -y @smithery/cli@latest install @NosytLabs/presearch-search-api-mcp --client claude ``` --- ## 🧪 Development ### Install Dependencies ```bash npm install ``` ### Run Tests ```bash # Run real API tests (requires .env with API key) npm run test:search # Run mock tool tests npm run test:tools ``` ### Build & Lint ```bash npm run lint npm run format ``` --- ## 🤝 Contributing Contributions are welcome! Please feel free to submit a Pull Request. <div align="center"> | Step | Action | |:-----|:-------| | 1 | Fork the repository | | 2 | Create your feature branch (`git checkout -b feature/AmazingFeature`) | | 3 | Commit your changes (`git commit -m 'Add some AmazingFeature'`) | | 4 | Push to the branch (`git push origin feature/AmazingFeature`) | | 5 | Open a Pull Request | </div> --- <div align="center"> _{Built with ❤️ for the decentralized web.} </div> ## 📜 License MIT © [Presearch MCP Team](https://github.com/NosytLabs)