Which integrations are available for this server?

Provides web search capabilities using Brave Search engine, with results including title, URL, and snippet. Enables web search via DuckDuckGo for privacy-respecting results. Allows searching GitHub for code repositories and related content. Connects to a self-hosted SearXNG instance for aggregated metasearch across multiple engines, with full privacy control. Supports web search using Sogou, a Chinese search engine. Integrates Sogou WeChat search to find WeChat articles.

How do I use @caikiji/mcp-search?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@@caikiji/mcp-search search for recent developments in AI" 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.

@caikiji/mcp-search

by caikiji

Overview Schema Related Servers Score Discussions

JavaScript

Remote

✨ Features

🌐 Web search — aggregates results from multiple engines (Brave, DuckDuckGo, Bing News, etc.)
📄 Page fetch — read full page content from search results as Markdown
🔍 Engine discovery — info to see engine counts; info scope: "engines" to list all
🎯 Token efficient — compact mode, smart snippet truncation, deduplicated results
🔒 Self-hosted — works with your own SearXNG instance, full privacy control

Installation

npm install -g @caikiji/mcp-search

Configuration

Environment variables

Variable	Default	Description
`SEARCH_URL`	—	Your SearXNG instance URL (required)
`SEARCH_AUTH`	—	Basic auth credentials (`user:pass`). Leave empty if your instance is public.
`SEARCH_TIMEOUT`	`15000`	SearXNG API timeout (ms)
`SEARCH_FETCH_TIMEOUT`	`15000`	Page fetch timeout (ms)
`SEARCH_DEFAULT_COUNT`	`10`	Default results per search (1–50)
`SEARCH_MAX_LENGTH`	`8000`	Max chars for `result` content (500–50000)
`SEARCH_SNIPPET_LENGTH`	`300`	Max chars for result snippet (truncated at word boundary)
`SEARCH_FETCH_UA`	Chrome UA	User-Agent for page fetching
`SEARCH_RECOMMENDED_ENGINES`	—	Semicolon-separated recommended engines, optionally with `engine`\|`description`. Shown in `info` output. See config example below.

MCP Client config

{
  "mcpServers": {
    "search": {
      "command": "npx",
      "args": ["-y", "@caikiji/mcp-search"],
      "env": {
        "SEARCH_URL": "https://search.example.com",
        "SEARCH_AUTH": "user:password",
        "SEARCH_RECOMMENDED_ENGINES": "duckduckgo|general purpose, privacy;bing|images, news;sogou|Chinese web;sogou wechat|WeChat articles;stackoverflow|programming Q&A;github|code repositories"
      }
    }
  }
}

Tools

Search

Tool	Arguments	Description
`query`	`query`, `[language]`, `[categories]`, `[time_range]`, `[engines]`, `[pageno]`, `[count]`, `[format]`	Web search with dedup, snippet, and source info. `format: "compact"` for minimal token usage.
`result`	`url`, `[max_length]`	Fetch a URL from results and return as Markdown. Works on most blogs/docs sites.
`info`	`[scope]`	Instance info. Default: counts, categories, settings. `scope: "engines"` for full list with categories.

`query` parameters

Param	Description
`query`	Search query (required)
`site`	Restrict to domain, e.g. `github.com`. Prepends `site:` to the query.
`language`	Language code (`zh-CN`, `en-US`, `auto`). Default: `auto`
`categories`	Comma-separated: `general`, `news`, `images`, `video`, `music`, `it`, `science`, `files`, `social media`
`time_range`	`day`, `week`, `month`, `year`
`engines`	Comma-separated engine names. Use `info` to see available ones
`pageno`	Page number. Default: `1`
`count`	Results to return (1–50). Default: `10`
`format`	`full` (title+URL+snippet) or `compact` (title+URL only, minimal tokens). Default: `full`
`depth`	When `true`, forces `count=1` and fetches the top result's full content via reader mode. Saves the query→pick→fetch round trip. Default: `false`

Output format

Full mode:

5 results | Sources: duckduckgo,brave
[1] Title
URL: https://...
Src: duckduckgo

Snippet text here (truncated to 300 chars)...

[2] Next Title
URL: https://...
Src: google +2

Compact mode:

5 results | Sources: duckduckgo
[1] Title — https://...
[2] Next Title — https://...

This server cannot be installed

license - permissive license

quality - not tested

maintenance

How are these scores calculated?

Resources

Need Help?

Related Servers

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Latest Blog Posts

Lightport: Open-Sourcing Glama's AI Gateway
By punkpeye on April 27, 2026.
open source
OpenAI
Tool Definition Quality Score (TDQS)
By punkpeye on April 3, 2026.
mcp
The Hackers Who Tracked My Sleep Cycle
By punkpeye on March 26, 2026.
security

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/caikiji/mcp-search'

If you have feedback or need assistance with the MCP directory API, please join our Discord server