Profitlee-MCP
ProfitLee MCP is an MCP server for analyzing ecommerce and marketplace profitability. It helps users calculate net profit, profit margin, ROI, breakeven price, and fee-adjusted outcomes from product cost, selling price, shipping, ads, platform fees, and tax inputs. It also supports reusable profit scenarios so users can compare pricing and cost assumptions more easily.
README
profitlee-mcp
An MCP server for Profitlee — compute country-accurate Amazon FBA/FBM and TikTok Shop profit margins, and manage saved scenarios, from any MCP client (Claude Desktop, Claude Code, Cursor, …).
calculate_profit is free and needs no token. The scenario tools require a Profitlee Pro API token.
MCP registry name:
io.github.AronLEEdev/profitlee-mcp
Quick start
Add to your MCP client config:
{
"mcpServers": {
"profitlee": {
"command": "npx",
"args": ["-y", "profitlee-mcp"],
"env": {
"PROFITLEE_API_TOKEN": "eck_live_xxx"
}
}
}
}
PROFITLEE_API_TOKEN is optional — omit the whole env block to use calculate_profit only. Create a token on your Profitlee account page to unlock the scenario tools.
Requires Node.js 20+.
Tools
| Tool | Auth | Description |
|---|---|---|
calculate_profit |
none | Full per-unit cost stack, gross/net margin, and monthly P&L. |
list_scenarios |
Pro token | List your saved scenarios. |
get_scenario |
Pro token | Read one scenario (inputs + outputs) by id. |
save_scenario |
Pro token | Save a named scenario from calculator inputs. |
update_scenario |
Pro token | Rename and/or replace a scenario's inputs. |
delete_scenario |
Pro token | Delete a scenario by id. |
calculate_profit inputs
Pick a platform + mode, give the product's physical and cost details, and Profitlee folds every fee into a single net margin. Rates are 0–1 decimals (e.g. 0.15 = 15%). US uses inches + pounds; DE/JP use cm + kg.
| Field | Notes |
|---|---|
platform |
amazon (default) or tiktok_shop. |
region |
us, de, or jp. |
mode |
amazon: fba | fbm. tiktok_shop: fbt | self_fulfilled. |
L, W, H, weight |
Dimensions + unit weight. |
fob, headShip, duty |
Unit cost, inbound freight/unit, import duty/unit. |
price |
Selling price (gross; VAT-inclusive for DE/JP). |
ppcAcos, returnRate |
Ad ACoS and return rate (0–1). |
monthlyVolume |
Units/month (scales the P&L). |
referralPct |
Referral fee (0–1). Preferred over referralCategory. |
isApparel |
Affects some fees. |
| mode-specific | FBA: inboundOption, storageMonths, storageSeason. FBM / TikTok self-fulfilled: outboundShipPerUnit, pickPackPerUnit, monthly3plStorage. TikTok FBT: storageMonthsPastFree. |
The Profitlee API is the source of truth for validation — incomplete or out-of-range inputs come back as a clear error listing the offending fields. Full field reference: https://profitlee.com/docs/api.
Environment variables
| Var | Required | Default | Purpose |
|---|---|---|---|
PROFITLEE_API_TOKEN |
No | — | Pro token (eck_live_…); needed only for the scenario tools. |
PROFITLEE_BASE_URL |
No | https://profitlee.com |
Override the API origin (testing). |
How it works
The server is a thin wrapper over Profitlee's public HTTP API:
calculate_profit→POST /api/v1/calculate(public, no token).- scenario tools →
/api/v1/scenarios*(require the Pro token; the server fails fast with a clear message if it's missing).
No fee logic is reimplemented here, so results always match the live Profitlee calculator and current fee tables.
Development
npm install
npm test # vitest (27 tests)
npm run build # tsc -> dist/
npm run dev # run from source with tsx
Releasing
Maintainers: see PUBLISHING.md for npm publish + MCP registry steps. The registry manifest lives in server.json.
License
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。