Provides direct AI-powered answers with citations via the KEIRO API.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@KEIRO MCP Server Search for the latest AI news"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
This is the official Model Context Protocol (MCP) server for the KEIRO API. It enables AI assistants like Claude to search the web, conduct deep research, extract web content, and get AI-powered answers—all in real-time.
Related MCP server: Tavily MCP Server
# Clone the repository
git clone https://github.com/keiro-labs/keiro-mcp.git
cd keiro-mcp
# Install dependencies
npm install
# Build
npm run buildSign up at keiro.dev and get your API key from the dashboard.
Add to your Claude Desktop config file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"keiro": {
"command": "node",
"args": ["/absolute/path/to/keiro-mcp/dist/index.js"],
"env": {
"KEIRO_API_KEY": "your_api_key_here"
}
}
}
}The KEIRO tools will now be available in Claude.
Tool | Description |
| Fast web search with AI-optimized results |
| Enhanced search with metadata and related queries |
| Deep research with summary and key points |
| Comprehensive research with analysis and sources |
| Direct AI-powered answers with citations |
| Extract content from any web page |
Once configured, you can ask Claude things like:
"Search for the latest developments in quantum computing"
"Research the impact of AI on healthcare in 2024"
"What are the key features of the new iPhone?"
"Extract the main content from https://example.com/article"
Claude will use the appropriate KEIRO tool to fetch real-time information.
Variable | Required | Description |
| Yes | Your KEIRO API key |
| No | Custom API URL (default: production) |
# Watch mode
npm run dev
# Type check
npm run typecheck
# Build
npm run buildThis MCP server uses the following KEIRO API endpoints:
POST /search - Basic search
POST /search-pro - Advanced search
POST /research - Topic research
POST /research-pro - Deep research
POST /answer - AI answers
POST /web-crawler - Page extraction
Full API documentation: keiro.dev/docs
MIT © KEIRO
Get a direct, AI-powered answer to any question with inline source citations. Best for factual questions requiring authoritative responses.
Inputs: query (required).
Extract and parse content from any web page. Returns clean, structured text optimized for AI processing. Handles JavaScript-rendered pages.
Inputs: url (required).
Conduct comprehensive research on any topic. Returns a detailed summary, key points, timeline, and cited sources. Perfect for in-depth analysis.
Inputs: query (required).
Advanced research with detailed analysis, multiple perspectives, data visualizations, and comprehensive source evaluation. For complex research tasks.
Inputs: query (required).
Search the web using KEIRO's semantic search engine. Returns relevant results with titles, URLs, and content snippets optimized for AI consumption.
Inputs: query (required).
Advanced search with enhanced results, metadata, related queries, and deeper content extraction. Best for comprehensive information gathering.
Inputs: query (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 6 tools
Multiple tools serve search and research purposes, with base and pro variants that overlap in functionality. The descriptions clarify output differences (e.g., search returns links, research returns summaries), reducing confusion, but an agent may still hesitate between search and research for a given task.
All tools follow the keiro_ prefix and use a consistent pattern of action terms (search, research, answer, crawl) with a uniform _pro suffix for advanced variants. Naming is highly predictable and easy to navigate.
Six tools is well within the ideal 3-15 range. The set covers basic and advanced search, basic and advanced research, direct answering, and crawling, which is well-scoped for a web intelligence server without unnecessary redundancy.
The tool surface covers the main stages of online information gathering: searching, deep research, direct Q&A, and page extraction. Minor gaps exist, such as lacking a tool for iterating through paginated search results, but these can be worked around with existing capabilities.
Web scraping, search, crawling, structured extraction and research for AI agents.
Give AI assistants access to real-time data. Search the web, compare flights, find hotels, and more.
Agent-native search engine with live web research optimized for AI agents.
Provides AI assistants with access to Seltz's powerful Web Search capabilities.
