SearXNG と統合し、プライバシー重視のメタ検索機能を提供する MCP サーバー実装。
メタ検索: 複数の検索エンジンの結果を結合します
プライバシー重視：追跡なし、ユーザープロファイリングなし
複数のカテゴリ: 一般、ニュース、科学、ファイル、画像、ビデオなどをサポート
言語サポート: 特定の言語またはすべての言語で検索
時間範囲フィルタリング: 日、週、月、年で結果をフィルタリングします
セーフサーチ: 3段階のセーフサーチフィルタリング
フォールバックサポート: 信頼性のための複数の SearXNG インスタンス
Related MCP server: SearXNG MCP Server
npm install -g @kevinwatt/mcp-server-searxngmcp-server-searxngDiveデスクトップで「+ MCPサーバーを追加」をクリックします。
次の設定をコピーして貼り付けます。
{
"mcpServers": {
"searxng": {
"command": "npx",
"args": [
"-y",
"@kevinwatt/mcp-server-searxng"
]
}
}
}「保存」をクリックしてMCPサーバーをインストールします
ウェブ検索
複数のエンジンでメタ検索を実行する
入力:
query （文字列）: 検索用語
page （数値、オプション）：ページ番号（デフォルト：1）
language (文字列、オプション): 言語コード (例: 'en'、'all'、デフォルト: 'all')
categories （配列、オプション）：検索カテゴリ（デフォルト：['general']）
利用可能: 「一般」、「ニュース」、「科学」、「ファイル」、「画像」、「動画」、「音楽」、「ソーシャルメディア」、「IT」
time_range (文字列、オプション): 時間フィルター (日/週/月/年)
safesearch (数値、オプション): セーフサーチレベル (0: なし、1: 中、2: 厳密、デフォルト: 1)
git clone https://github.com/kevinwatt/mcp-server-searxng.git
cd mcp-server-searxng
npm install
npm run build
npm startこのMCPサーバーはMITライセンスに基づいてライセンスされています。詳細はLICENSEファイルをご覧ください。
ローカルのSearXNGインスタンスが稼働している必要があります。設定方法は次のとおりです。
# Create config directory
mkdir -p searxng
# Create config file
tee searxng/settings.yml << EOF
use_default_settings: true
server:
bind_address: "0.0.0.0"
secret_key: "CHANGE_THIS_TO_SOMETHING_SECURE" # Generate a random key
port: 8080
search:
safe_search: 0
formats:
- html
- json
engines:
- name: google
engine: google
shortcut: g
- name: duckduckgo
engine: duckduckgo
shortcut: d
- name: bing
engine: bing
shortcut: b
server.limiter: false
EOF
# Start container
docker run -d \
--name searxng \
-p 8080:8080 \
-v "$(pwd)/searxng:/etc/searxng" \
searxng/searxng# Test JSON API with curl
curl -v 'http://localhost:8080/search?q=test&format=json'
# Or visit in browser
http://localhost:8080/search?q=test# Stop container
docker stop searxng
# Remove container
docker rm searxng
# View container logs
docker logs searxng
# Enable auto-start on boot
docker update --restart always searxng--restart alwaysフラグにより、次のことが保証されます。
Dockerデーモンが起動するとコンテナが自動的に起動します
コンテナはクラッシュすると自動的に再起動します
コンテナは、ユーザーが明示的に停止しない限り、停止すると自動的に再起動します。
searxng/settings.yml次のように編集します。
検索エンジンリストを変更する
セキュリティ設定を調整する
UI言語を設定する
API制限の変更
詳細な設定オプションについては、 SearXNGドキュメントを参照してください。
SEARXNG_INSTANCES : SearXNGインスタンスURLのカンマ区切りリスト デフォルト: http://localhost:8080
SEARXNG_USER_AGENT : リクエストのカスタム User-Agent ヘッダー デフォルト: MCP-SearXNG/1.0
NODE_TLS_REJECT_UNAUTHORIZED : SSL証明書の検証をバイパスするには '0' に設定します（自己署名証明書を使用した開発用）。デフォルト: 未定義（SSL検証が有効）
すべてのオプションを含む構成例:
{
"mcpServers": {
"searxng": {
"name": "searxng",
"command": "npx",
"args": [
"-y",
"@kevinwatt/mcp-server-searxng"
],
"env": {
"SEARXNG_INSTANCES": "http://localhost:8080,https://searx.example.com",
"SEARXNG_USER_AGENT": "CustomBot/1.0",
"NODE_TLS_REJECT_UNAUTHORIZED": "0"
}
}
}
}⚠️ 警告: 実稼働環境では SSL 証明書の検証を無効にすることはお勧めしません。
Performs a web search using SearXNG, ideal for general queries, news, articles and online content. Supports multiple search categories, languages, time ranges and safe search filtering. Returns relevant results from multiple search engines combined.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | Search query | |
| page | No | Page number (default 1) | |
| language | No | Search language code (e.g. 'en', 'zh', 'jp', 'all') | all |
| categories | No | ||
| time_range | No | all_time | |
| safesearch | No | 0: None, 1: Moderate, 2: Strict |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full behavioral burden. It discloses that results come from multiple search engines combined and that filtering options are supported, but it does not mention limitations, rate limits, result count, or whether the operation is read-only. This is adequate but not deeply transparent.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Three sentences, each earning its place: purpose, supported controls, and output behavior. The most important information is front-loaded, with no filler or repeated schema content.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a six-parameter tool with no output schema and no annotations, the description covers purpose and high-level behavior but not result format, pagination behavior, or edge cases like empty results. It is sufficient for basic invocation but leaves some context to be inferred.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 67%, and the schema already documents query, page, language, and safesearch values. The description adds a high-level statement that categories, languages, time ranges, and safe search are supported, but it does not add details beyond the schema. The generic references are not enough to meaningfully augment the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description opens with a specific verb and resource: 'Performs a web search using SearXNG', and enumerates concrete use cases such as general queries, news, articles, and online content. It also distinguishes itself from a plain search by noting results come from multiple search engines combined. With no sibling tools to contrast against, this is maximum clarity.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description states the tool is 'ideal for general queries, news, articles and online content', which provides clear usage context. It does not state explicit when-not-to-use conditions or alternatives, but that is reasonable since no sibling tools are listed.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
web_searchScored across 1 tool
With only one tool, there is no possibility of ambiguity or overlap between tools. The tool's purpose is clearly defined as performing web searches using SearXNG, making it distinct by default.
A single tool inherently has perfect naming consistency, as there are no other tools to compare it against. The name 'web_search' follows a clear verb_noun pattern and is appropriate for its function.
One tool is too few for a server named 'SearXNG MCP Server', which suggests a broader scope for search functionality. While the tool covers general web searches, the lack of additional tools (e.g., for image search, news-specific searches, or configuration) makes the set feel thin and underdeveloped for the domain.
The tool surface is severely incomplete for a search server. It only provides general web search, missing obvious extensions like image search, video search, news search, or advanced filtering tools that would be expected in a comprehensive search interface, leading to significant gaps in functionality.
Search the web, images, videos, news, and local businesses with robust filters, freshness controls…
Web search with quality scores; web pages are read on your computer by the local server.
Google, Bing, DuckDuckGo and Google Maps search results, free without an API key.
Search privacy tools, guides, stacks and glossary, find app replacements, run privacy checks.
