com.noteboxd由 noteboxd.com 發布官方發布者v0.1.4更新於 Oct 27, 2026
Fragrance & perfume MCP: search fragrances, notes, accords, brands, perfumers, reviews & charts.
Noteboxd 香水資料庫 MCP：讓助理搜尋香水、香調、品牌、調香師、評論與排行榜，並讀取使用者自己的收藏。
Web executable，透過 Streamable HTTP。 遠端服務在工作區中設定後即可從網頁執行環境執行。
把這段提示詞貼上到 Claude Code、Codex、Cursor 等代理程式，或 SourceWeft 對話中。代理程式會讀取該伺服器的安裝指南，先向你展示發布者、端點和登入方式，經你同意後再把它新增到你的客戶端。金鑰不會寫進任何命令。
閱讀 https://sourceweft.com/mcp/com.noteboxd%2Fmcp/install.md，並按其中說明把這個 MCP 伺服器接入我的客戶端。修改任何設定之前，先告訴我它的發布者、端點和登入方式，等我確認。不要把 API 金鑰寫進命令或我會提交的檔案；問我金鑰應從哪裡取得。把它新增到你客戶端的 mcpServers 設定中。
{
"mcpServers": {
"fragrance-perfume-noteboxd": {
"type": "http",
"url": "https://mcp.noteboxd.com/mcp"
}
}
}A Model Context Protocol server (and API) for fragrance and perfume data. Connect Claude, ChatGPT, Cursor, or any MCP client to the Noteboxd fragrance encyclopedia: fragrances, notes, accords, brands, perfumers, community reviews, and charts.
Once connected, you can ask things like:
The assistant answers from Noteboxd's data instead of guessing.
Every client uses the same command (npx -y @noteboxd/mcp) with your key in NOTEBOXD_API_KEY.
Add to claude_desktop_config.json:
Then set your real key. Or add the JSON block above to ~/.cursor/mcp.json.
Noteboxd also runs a hosted MCP server at https://mcp.noteboxd.com/mcp (Streamable HTTP, with
OAuth sign-in) if you would rather not run anything locally. This package is the local,
open-source option.
All tools are read-only and return JSON. Each call is authenticated with your key and counts against your Noteboxd developer plan (the free tier covers 50 calls/day).
You can also pass the key with --api-key nb_live_....
This is a thin local server. It speaks MCP to your client over stdio and forwards each tool call
to api.noteboxd.com/v1 using your key. It stores no data and holds no credentials beyond the key
you provide.
See examples/ for runnable code: curl, TypeScript, and Python.
The server is free and open source. API usage follows your Noteboxd developer plan: a free tier (50 calls/day), then pay-as-you-go per call. Most reads are $0.01; the AI tools cost a few cents. See developers.noteboxd.com/#pricing.
來源：README.md，提交 18200d2
Open the Noteboxd app UI in ChatGPT. Call with no arguments to show the home screen (trending fragrances), with `query` to run a fragrance search inside the app, or with `fragrance` (slug or UUID) to show a fragrance profile. The app renders the returned result; keep the chat reply to one short line.
queryfragranceUniversal search across all Noteboxd entities — fragrances, brands, notes, perfumers, and collections — in a single call. Best for quick lookups and vague queries. Examples: - "Jacques Polge" → returns the perfumer - "oud" → returns the note, plus top oud fragrances - "Bleu" → returns Bleu de Chanel, Bleu de Dior, etc. See also: `fragrance_search` for advanced fragrance-only filtering (by note, score, concentration) · `trending` for what's popular right now.
query必填response_formatWhat's trending on Noteboxd right now — top search queries, highest-rated fragrances, and most popular notes. No parameters needed. See also: `charts_list` for full ranking leaderboards · `fragrance_search` to explore specific categories.
response_formatFind fragrances with similar accord profiles to a given fragrance. Returns ranked alternatives based on shared accords and character. See also: `fragrance_get` for the source fragrance's full profile · `fragrance_search` with note/accord filters for manual exploration.
id必填response_formatList all available Noteboxd chart rankings (e.g. top-rated, trending, top by brand). Returns chart names and summary stats. See also: `chart_get` with a chart slug to see the full leaderboard · `trending` for a quick snapshot of what's hot.
response_formatGet a specific chart leaderboard by slug (e.g. 'top-rated', 'trending', 'top-rated-brand-chanel'). Returns ranked fragrances with scores and movement. See also: `charts_list` to see all available charts · `fragrance_get` for the full profile of any fragrance on the chart.
slug必填response_formatBrowse fragrance notes — filter by family (e.g. 'woody', 'floral', 'citrus') and sort by popularity or name. Returns note names, families, colours, and fragrance counts. See also: `note_explore` for a single note's full profile and top fragrances · `fragrance_search` with the `note` filter to find fragrances featuring a specific note.
sortlimitfamilyresponse_formatGet personalized fragrance recommendations based on your taste profile. Optional natural-language prompt to guide results (e.g. 'something fresh for summer'). Requires Pro. See also: `ai_wear_today` for what to wear right now · `user_get_cabinet` for your collection.
limitpromptresponse_formatWhat should I wear today? Picks from your owned cabinet (CURRENT status only) based on optional prompt/weather. Returns nothing if your cabinet is empty. Requires Pro. See also: `ai_recommend` for discovery beyond your cabinet · `user_get_cabinet` for your full collection.
promptweatherresponse_formatGenerate a curated fragrance collection from a natural-language prompt (e.g. 'office-safe designer scents under $100', 'niche oud masterpieces'). Returns 5–25 grounded picks with rationale. Requires Pro. See also: `ai_recommend` for personalized picks · `fragrance_search` for manual filtering.
sizeprompt必填response_formatSearch and filter the Noteboxd fragrance encyclopedia. Requires a query or at least one filter — the full catalogue cannot be listed. Tip: The `q` parameter searches fragrance names OR brand names independently. For best results with a known brand, use the `brand` slug filter instead of including the brand name in `q`. Examples: - "Find woody fragrances above 8/10" → accord='woody' + minScore=8 - "Creed fragrances" → brand='creed' - "Bleu de Chanel" → q='Bleu' + brand='chanel' See also: `fragrance_get` for a single fragrance's full profile · `search` for universal cross-entity lookup (brands, notes, perfumers, collections in one call).
notepagesortbrandlimitqueryaccordgenderminScoreconcentrationresponse_formatGet the full profile for a single fragrance by UUID or slug (e.g. 'tom-ford/oud-wood'). Returns notes pyramid, accords, FOTW classification, perfumer, community score, and more. See also: `fragrance_similar` for fragrances with similar accord profiles · `fragrance_get_reviews` for community reviews · `fragrance_get_wearing_stats` for when people wear it.
idslugresponse_formatGet paginated community reviews for a fragrance. Sort by helpfulness, recency, or score. See also: `fragrance_get` for the fragrance's full profile · `fragrance_get_wearing_stats` for aggregate wearing data.
id必填sortlimitcursorresponse_formatCommunity wearing counts for a fragrance. Free tier sees the last 30 days; Pro also sees today and the last 7 days. See also: `fragrance_get` for full profile · `fragrance_similar` for alternatives with similar profiles.
id必填response_formatGet a brand profile (name, country, founded, description) and its top fragrances. Use the brand slug, e.g. 'maison-francis-kurkdjian'. See also: `fragrance_search` with the `brand` filter to get the full paginated catalogue for this brand · `perfumer_get` if the brand has a known house perfumer.
slug必填response_formatGet a perfumer's profile (bio, country, house affiliation) and their top fragrances by community score. See also: `fragrance_search` to find more fragrances by this perfumer · `brand_get` if you want the brand they work for.
slug必填response_formatGet a note's profile (family, description, origin, extraction method) and top fragrances featuring it, sorted by community score. Use the note slug, e.g. 'oud', 'bergamot', 'iris'. See also: `notes_list` to browse all notes by family/popularity · `fragrance_search` with the `note` filter to find fragrances featuring this note.
note必填response_formatCompare two fragrances head-to-head — shared vs. distinct notes/accords/families, community and critic scores, audience, and an AI-written synthesis (shared character, differences, who each suits). Public, cached 7 days per pair. See also: `fragrance_get` for full profiles · `fragrance_similar` for alternatives.
fragrance_a必填fragrance_b必填response_formatGet a concise AI-generated digest summarizing a fragrance's community reviews — key themes, consensus, and standout opinions. Public, cached per fragrance. See also: `fragrance_get_reviews` for individual reviews · `fragrance_get` for full profile.
id必填response_formatGet your fragrance cabinet — fragrances you own, have tried, or finished. Requires a Pro account. See also: `fragrance_get` for full details on any fragrance in your cabinet · `fragrance_similar` to find alternatives to your favourites · `user_get_wishlist` for fragrances you want to try.
sortlimitcursorstatusresponse_formatGet your fragrance wishlist. Requires a Pro account. See also: `fragrance_get` for full details on any wishlist fragrance · `user_get_cabinet` for fragrances you already own.
limitcursorresponse_formatGet your collections (fragrance lists). Optionally pass a collectionId to see its items. Requires a Pro account. See also: `fragrance_get` for full details on any fragrance in a collection.
collectionIdresponse_formatGet reviews you've written, or reviews you found helpful. Requires a Pro account. See also: `fragrance_get_reviews` for community reviews on a specific fragrance.
kindlimitcursorresponse_formatGet your wearing check-in history over time. Requires a Pro account. See also: `fragrance_get_wearing_stats` for aggregate community wearing data on a specific fragrance.
limitcursorresponse_formatGet a user's public taste profile — their top notes, accords, and families derived from their collection and reviews. No LLM call, deterministic. See also: `user_taste_compare` for compatibility between two users.
username必填response_formatGet your AI-generated taste profile summary — a natural-language description of your fragrance preferences based on your cabinet, reviews, and likes. Requires Pro. See also: `user_taste_public` for the deterministic breakdown · `user_taste_compare` for compatibility.
response_formatCompare your taste with another user — compatibility score, shared and divergent notes/accords/families, and an optional AI overview. Requires Pro for the overview. See also: `user_taste_public` for a user's top preferences · `user_taste_summary` for your own summary.
with_overviewtarget_user_id必填response_format已滿足適用檢查項的 100%
該評分衡量此收錄條目的完整程度，依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價，也不表示伺服器安全或目前可用：可連通性是帶日期的檢查結果，評分也不包含安裝量或使用者評價。
NOTEBOXD_API_KEY=nb_live_your_key npx -y @noteboxd/mcp{ "mcpServers": { "noteboxd": { "command": "npx", "args": ["-y", "@noteboxd/mcp"], "env": { "NOTEBOXD_API_KEY": "nb_live_your_key" } } }}{ "servers": { "noteboxd": { "command": "npx", "args": ["-y", "@noteboxd/mcp"], "env": { "NOTEBOXD_API_KEY": "nb_live_your_key" } } }}AI client --stdio--> noteboxd-mcp --HTTPS + your key--> api.noteboxd.com/v1