AI 어시스턴트를 위한 npm 패키지 트렌드 데이터 모든 npm 패키지의 주간 다운로드 수를 추적하세요. 패키지 다운로드 트렌드는 개발자들이 어떤 JavaScript 라이브러리와 프레임워크를 채택하고 있는지, 어떤 것이 입지를 잃고 있는지, 생태계가 어디로 향하고 있는지 보여줍니다.
전체 문서 및 라이브 데모: https://trendsmcp.ai/npm-trends
**Trends MCP**의 일부 – 12개 이상의 소스에서 실시간 트렌드 데이터를 제공하는 MCP 서버입니다. 메인 저장소: https://github.com/trendsmcp/trends-mcp
1단계: **trendsmcp.ai**에서 무료 API 키를 받으세요. 하루 100회 요청 가능, 신용카드 불필요.
2단계: AI 클라이언트에 추가하세요 (YOUR_API_KEY를 교체):
Cursor / Windsurf / Cline (~/.cursor/mcp.json 또는 동등 파일)
{
"mcpServers": {
"trends-mcp": {
"url": "https://api.trendsmcp.ai/mcp",
"transport": "http",
"headers": { "Authorization": "Bearer YOUR_API_KEY" }
}
}
}VS Code / GitHub Copilot (.vscode/mcp.json)
{
"servers": {
"trends-mcp": {
"type": "http",
"url": "https://api.trendsmcp.ai/mcp",
"headers": { "Authorization": "Bearer YOUR_API_KEY" }
}
}
}Claude Desktop (claude_desktop_config.json)
사용자 → 설정 → 개발자 → 구성 편집 — mcpServers 내부에 추가
{
"mcpServers": {
"trends-mcp": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://api.trendsmcp.ai/mcp",
"--header",
"Authorization:${AUTH_HEADER}"
],
"env": {
"AUTH_HEADER": "Bearer YOUR_API_KEY"
}
}
}
}Claude.ai (브라우저) 설정 -> 커넥터 -> 사용자 정의 커넥터 추가:
https://api.trendsmcp.ai/mcpRelated MCP server: npm-trends-mcp
연결 후 AI에게 물어보세요:
get_trends(keyword='react', source='npm', data_mode='weekly')도구 | 기능 설명 |
| 이 소스의 키워드에 대한 시계열 데이터 |
| 1주, 1개월, 3개월, 6개월, 1년 기간별 성장률 |
| 현재 이 소스에서 트렌딩 중인 항목 |
| 볼륨 기준 상위 주제 순위 |
모든 npm 패키지의 주간 다운로드 수를 0-100 척도로 정규화한 값과 원시 다운로드 볼륨, 표준 기간(7일, 1개월, 3개월, 1년)에 대한 성장률, 최대 5년까지의 과거 시계열을 반환합니다.
정확한 npm 패키지 이름을 사용하세요. 예: 'react', 'lodash', 'express', '@anthropic-ai/sdk'. 스코프 패키지는 @org/package 형식을 사용합니다.
네. 쉼표로 구분된 패키지 이름으로 get_growth를 사용하거나 각 패키지에 대해 get_trends를 호출하여 정규화된 시계열을 비교하세요. React vs Vue vs Svelte 같은 프레임워크 비교에 유용합니다.
npm 다운로드 트렌드는 개발자 생태계 채택의 선행 지표입니다. 다운로드가 빠르게 증가하는 라이브러리는 종종 기반 플랫폼이나 프레임워크의 상업적 채택을 앞서서 나타냅니다.
원시 주간 다운로드 수는 패키지 자체의 과거 최고치를 기준으로 0-100 척도로 정규화됩니다. 이를 통해 절대 다운로드 볼륨이 크게 다른 패키지 간의 채택 속도를 비교할 수 있습니다.
Trends MCP는 하나의 연결로 12개 이상의 소스를 다룹니다: Google 검색, YouTube, TikTok, Reddit, Amazon, Wikipedia, 뉴스 감성, 웹 트래픽, 앱 다운로드, Steam, npm 등.
전체 보기: https://trendsmcp.ai/data-sources
동일한 API 키가 Python에서 직접 작동합니다. MCP 호스트가 필요 없습니다.
pip install npm-trends-mcpimport os
from npm_trends_mcp import TrendsMcpClient, SOURCE
client = TrendsMcpClient(api_key=os.environ["TRENDSMCP_API_KEY"])
series = client.get_trends(source=SOURCE, keyword="your keyword")
growth = client.get_growth(source=SOURCE, keyword="your keyword", percent_growth=["1M", "3M", "12M"])
top = client.get_top_trends(type="Npm", limit=10)MIT © Trends MCP
Point-to-point growth for a keyword on one or more sources. Each window is a preset string (12M, 3M, YTD, and the other listed periods). Values are on a 0-100 scale, plus absolute volume when available. Prefer this over get_time_series for growth questions. app downloads and app rankings are keyword sources (Android bundle ID). They are not the App Store / Google Play live boards on get_top_trends. If the request is rate limited or the monthly quota is used up, tell the user their plan limit is reached.
| Name | Required | Description | Default |
|---|---|---|---|
| source | Yes | One source, or comma-separated sources (e.g. 'amazon, tiktok, youtube'). Valid: 'google search', 'google images', 'google news', 'google shopping', 'youtube', 'wikipedia', 'tiktok', 'reddit', 'amazon', 'news sentiment', 'news volume', 'npm', 'python', 'steam', 'app downloads', 'app rankings'. | |
| keyword | Yes | What to look up. The string format is required by source. Standard sources (google search, google images, google news, google shopping, youtube, wikipedia, tiktok, reddit, amazon, news sentiment, news volume): any name or phrase, e.g. 'nike'. npm: exact npmjs.com package name, case-sensitive. Right: 'react', '@babel/core'. Wrong: 'React', 'React.js'. python: exact PyPI project name. Right: 'pandas', 'requests'. Wrong: 'Pandas'. steam: game display name in plain English, not a Steam App ID. Right: 'Elden Ring', 'CS2'. First Steam store search result wins, so use an unambiguous name. app downloads and app rankings: Android bundle ID only (the id= value on Google Play). Right: 'com.openai.chatgpt', 'com.whatsapp'. Wrong: 'ChatGPT', 'WhatsApp', an iOS App Store ID, or a bundle ID that is not Android. Find it at play.google.com/store/apps/details?id=THIS_PART. If the request includes app downloads or app rankings with other sources, keyword must still be the Android bundle ID. | |
| percent_growth | No | Growth windows. Default if omitted: ['12M']. Each item must be a preset string: '7D', '1W', '14D', '2W', '30D', '1M', '2M', '3M', '6M', '9M', '12M', '1Y', '18M', '24M', '2Y', '36M', '3Y', '48M', '4Y', '60M', '5Y', 'MTD', 'QTD', 'YTD'. Every preset is a two-date comparison. |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Beyond the readOnlyHint, openWorldHint, and idempotentHint annotations, the description adds useful behavioral context: values are on a 0-100 scale with absolute volume when available, and the agent should tell the user their plan limit is reached on rate limiting or quota exhaustion. These details are not present in the structured fields and do not contradict the annotations.
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?
The description is compact and every sentence earns its place: core operation, output scale, sibling preference, app-source caveat, and rate-limit handling. It front-loads the primary purpose and avoids redundant phrasing.
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?
With a detailed input schema, an output schema, and safety annotations, the description fills the remaining gaps: tool selection, output scale, source disambiguation, and error behavior. It is complete enough for an agent to invoke the tool correctly without further inference.
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?
The input schema already documents all parameters thoroughly, including valid sources, keyword formatting rules per source, and every preset growth window. The description references preset windows and Android bundle IDs but does not add meaning beyond the schema, so a baseline of 3 is appropriate.
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 action and resource: 'Point-to-point growth for a keyword on one or more sources.' It also names sibling tools to disambiguate itself, stating it is preferred over get_time_series for growth questions and clarifying that app sources are not the live boards on get_top_trends.
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?
It gives an explicit selection rule: 'Prefer this over get_time_series for growth questions.' It also warns against confusing app downloads/app rankings with get_top_trends live boards, and specifies how to handle rate limits or quota exhaustion. This gives the agent clear when-to-use and when-not-to-use guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Full historical series for one keyword and one source (0-100 values, plus volume when available). Use for charting or custom math. Not for live 'what's trending now' boards (use get_top_trends). For most growth questions, use get_growth. If the request is rate limited or the monthly quota is used up, tell the user their plan limit is reached.
| Name | Required | Description | Default |
|---|---|---|---|
| source | Yes | Exactly one source per request. Valid: 'google search', 'google images', 'google news', 'google shopping', 'youtube', 'wikipedia', 'tiktok', 'reddit', 'amazon', 'news sentiment', 'news volume', 'npm', 'python', 'steam', 'app downloads', 'app rankings'. | |
| keyword | Yes | What to look up. The string format is required by source. Standard sources (google search, google images, google news, google shopping, youtube, wikipedia, tiktok, reddit, amazon, news sentiment, news volume): any name or phrase, e.g. 'tesla'. npm: exact npmjs.com package name, case-sensitive. Right: 'react', '@babel/core'. Wrong: 'React', 'React.js'. python: exact PyPI project name. Right: 'pandas', 'requests'. Wrong: 'Pandas'. steam: game display name in plain English, not a Steam App ID. Right: 'Elden Ring', 'CS2'. First Steam store search result wins, so use an unambiguous name. app downloads and app rankings: Android bundle ID only (the id= value on Google Play). Right: 'com.openai.chatgpt', 'com.whatsapp'. Wrong: 'ChatGPT', 'WhatsApp', an iOS App Store ID, or a bundle ID that is not Android. Find it at play.google.com/store/apps/details?id=THIS_PART. |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover readOnlyHint and idempotentHint. The description adds behavioral context beyond those hints: the 0-100 value range, volume availability, and rate-limit/quota handling instructions. This provides useful operational detail that annotations do not convey, though the core safety profile is already covered by hints.
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 and scope, usage routing, and error handling. Front-loaded with the core purpose; no repetition or fluff. The alternative tool routing is included efficiently.
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?
Given a two-parameter tool with a full output schema, the description covers purpose, usage scenarios, alternatives, and error handling. An agent has all necessary information to decide when to call it and what to expect, with no critical gaps.
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 description coverage is 100% and both parameters have detailed, source-specific descriptions (e.g., keyword format per source, valid source list). The tool description adds no additional parameter meaning beyond restating 'one keyword and one source', so the baseline for full schema coverage applies.
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 explicitly states it returns 'Full historical series for one keyword and one source' with a value range and volume detail. It distinguishes itself from siblings get_top_trends and get_growth by stating what it is not for (live trending, growth questions), making the tool's purpose unambiguous.
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?
Directly states when to use it: 'Use for charting or custom math' and when not to: 'Not for live trending (use get_top_trends)' and 'For most growth questions, use get_growth'. Also provides explicit handling for rate-limited or quota-exhausted requests, which is actionable guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Live top-trending board for exactly one feed type. No keyword. For 'Amazon Best Sellers by Category', 'Google Trends by Category', 'Top Websites', and 'Substack by Category', always pass category. Default sort is current rank. Use sort='rank_change' for climbers vs a prior snapshot (window 1d, 3d, 7d, 14d, or 30d). App Store Top Free, App Store Top Paid, and Google Play are live store boards, not keyword lookups. For an app's history use get_growth or get_time_series with source app downloads or app rankings and an Android bundle ID. Do not use get_time_series for live boards. If the request is rate limited or the monthly quota is used up, tell the user their plan limit is reached.
| Name | Required | Description | Default |
|---|---|---|---|
| sort | No | How to rank the board. 'rank' (default): current leaders. 'rank_change': biggest climbers vs a prior snapshot. Mover rows include rank, keyword, prev_rank, and rank_change. | rank |
| type | Yes | Exactly one live feed. Valid: 'Amazon Best Sellers Top Rated', 'Amazon Best Sellers by Category', 'App Store Top Free', 'App Store Top Paid', 'GitHub', 'Google News Top News', 'Google Play', 'Google Trends', 'Google Trends by Category', 'IMDb MOVIEmeter', 'Open Library Trending Books', 'Reddit Hot Posts', 'Reddit World News', 'Top Websites', 'Spotify Top Podcasts', 'Steam Most Played', 'Substack', 'Substack by Category', 'TikTok Trending Hashtags', 'TikTok Trending Searches', 'Wikipedia Trending', 'X (Twitter) Trending', 'YouTube Trending'. | |
| limit | No | Max rows to return. Default 25, min 1, max 200. | |
| offset | No | Rows to skip for pagination. Default 0. | |
| window | No | Lookback used only when sort is 'rank_change'. One of '1d', '3d', '7d', '14d', '30d'. Default '30d'. Short windows only work on daily feeds; weekly and monthly feeds return a note pointing to a longer window. | 30d |
| category | No | Pass this whenever type is 'Amazon Best Sellers by Category', 'Google Trends by Category', 'Top Websites', or 'Substack by Category'. Use the official name. Without it those feeds mix every board. Only omit on a first pull to learn the official names. |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Beyond readOnlyHint/idempotentHint, the description adds important behavioral context: rate limits and monthly quota handling, default sort behavior, rank-change snapshot windows, category omission behavior, and feed-type quirks. It discloses how the tool behaves in edge cases, which is exactly the kind of transparency that helps an agent.
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?
The description is dense but every sentence earns its place: purpose, category rule, sort semantics, alternative tool routing, and rate-limit behavior are all covered in a compact, front-loaded format. No filler or repetition.
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?
Given six parameters, sibling tools, and an output schema, the description covers everything an agent needs: valid feed types, category requirements, sort/window semantics, pagination defaults handled in schema, alternatives, and rate-limit response. Nothing essential is missing.
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 100%, so the baseline is 3. The description adds valuable non-obvious semantics: category is mandatory for four specific feed types and mixing occurs otherwise, and sort='rank_change' compares against a prior snapshot. This goes beyond the schema's own parameter descriptions.
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?
Description states a specific verb and resource: it returns a live top-trending board for exactly one feed type, and explicitly says 'No keyword.' This clearly distinguishes it from sibling tools like get_growth and get_time_series without needing to open the schema.
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 provides explicit when-to-use guidance: pass category for certain feeds, use sort='rank_change' for climbers, and use get_growth or get_time_series for app history. It even warns 'Do not use get_time_series for live boards,' giving clear exclusions and alternatives.
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.
v0.1.0get_growthget_time_seriesget_top_trendsScored across 3 tools
The tools are largely distinct: get_top_trends targets live boards, while get_growth and get_time_series both handle historical keyword data and could be confused. However, the descriptions explicitly separate 'point-to-point growth' from 'full historical series' and tell agents which to prefer, so the boundary is workable.
All three tools follow a clean get_<noun> pattern: get_growth, get_top_trends, get_time_series. There is no mixing of styles or inconsistent verbs, making the naming predictable and easy to navigate.
Three tools is a focused, sensible scope for a trends data server. Each tool covers a distinct core user need: current top boards, point-to-point growth, and full time-series data, so each earns its place.
The set covers the main trend-related workflows: live rankings, growth comparisons, and historical charting data. Minor gaps exist, such as no explicit endpoint for listing available feeds, categories, or discovering keywords, but agents can work around those with known values and guidance in the descriptions.
npm package download trends and adoption growth over time. Free key at trendsapi.ai
Package intelligence for AI agents across npm, PyPI, crates.io and deps.dev. No API keys.
Package intelligence for AI agents across npm, PyPI, crates.io and deps.dev. No API keys.
Provide AI-powered real-time analysis and intelligence on NPM packages, including security, depend…