🔍 Presearch MCP Server

🔐 Privacy-First AI Search Integration

Empower your AI with decentralized, uncensored web search capabilities through the Model Context Protocol (MCP).

🌟 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:

🛡️ 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

🔥 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

{
  "query": "climate change effects on coastal real estate",
  "breadth": 5,
  "depth": 2,
  "research_focus": "market"
}

Result: The agent autonomously performs a multi-step investigation:

Initial Search: Queries Presearch for broad market trends.
Analysis: Identifies key sub-topics (insurance rates, flood risk zones, property devaluation).
Deep Dive: Executes targeted sub-searches for each topic.
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

{
  "query": "iPhone 16 Pro camera specs official",
  "limit": 3
}

Result:

Search: Finds the top 3 most relevant pages (Apple Official, TechCrunch, The Verge).
Scrape: Immediately fetches the full HTML content of all 3 pages in parallel using a headless browser.
Output: Returns 35kb of clean text content.
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

{
  "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

🚫 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

💡 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

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)

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:

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

📚 Resources

The server exposes the following resources for configuration and debugging:

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

⚙️ Configuration

The server can be configured via environment variables or MCP settings.

Environment Variables

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`

JSON Configuration Schema

When using Smithery or an MCP client, the configuration object supports:

{
  "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

2. Run with npx

npx presearch-mcp-server

3. Deploy via Smithery

Use the button above or run:

npx -y @smithery/cli@latest install @NosytLabs/presearch-search-api-mcp --client claude

🧪 Development

Install Dependencies

npm install

Run Tests

# Run real API tests (requires .env with API key)
npm run test:search

# Run mock tool tests
npm run test:tools

Build & Lint

npm run lint
npm run format

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

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

_{Built with ❤️ for the decentralized web.}

📜 License

MIT © Presearch MCP Team

🔍 Presearch MCP Server

🔐 Privacy-First AI Search Integration

Empower your AI with decentralized, uncensored web search capabilities through the Model Context Protocol (MCP).

🌟 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:

🛡️ 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

🔥 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

{
  "query": "climate change effects on coastal real estate",
  "breadth": 5,
  "depth": 2,
  "research_focus": "market"
}

Result: The agent autonomously performs a multi-step investigation:

Initial Search: Queries Presearch for broad market trends.
Analysis: Identifies key sub-topics (insurance rates, flood risk zones, property devaluation).
Deep Dive: Executes targeted sub-searches for each topic.
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

{
  "query": "iPhone 16 Pro camera specs official",
  "limit": 3
}

Result:

Search: Finds the top 3 most relevant pages (Apple Official, TechCrunch, The Verge).
Scrape: Immediately fetches the full HTML content of all 3 pages in parallel using a headless browser.
Output: Returns 35kb of clean text content.
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

{
  "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?

Why Presearch Matters for AI

🚫 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

💡 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

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)

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:

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

📚 Resources

The server exposes the following resources for configuration and debugging:

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

⚙️ Configuration

The server can be configured via environment variables or MCP settings.

Environment Variables

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`

JSON Configuration Schema

When using Smithery or an MCP client, the configuration object supports:

{
  "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

2. Run with npx

npx presearch-mcp-server

3. Deploy via Smithery

Use the button above or run:

npx -y @smithery/cli@latest install @NosytLabs/presearch-search-api-mcp --client claude

🧪 Development

Install Dependencies

npm install

Run Tests

# Run real API tests (requires .env with API key)
npm run test:search

# Run mock tool tests
npm run test:tools

Build & Lint

npm run lint
npm run format

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

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

_{Built with ❤️ for the decentralized web.}