sparkpost-mcp
Enables sending transactional emails, managing templates, suppression lists, sending domains, webhooks, subaccounts, and retrieving account metrics via the SparkPost API.
README
sparkpost-mcp
MCP server exposing the SparkPost (EU) email API as 20 tools — account, templates, transactional sends, suppression, sending domains, webhooks, subaccounts, message events, deliverability metrics, and recipient lists.
Published on npm as @pcummings/sparkpost-mcp.
Install into an AI agent
You need a SPARKPOST_API_KEY (SparkPost EU dashboard → API Keys). For the US region, also set SPARKPOST_API_BASE=https://api.sparkpost.com/api/v1.
The commands below fetch the server from npm via
npx -y @pcummings/sparkpost-mcp. To run from a local clone instead, replacenpx -y @pcummings/sparkpost-mcpwithnpx tsx /absolute/path/to/index.ts.
Claude Code
claude mcp add sparkpost -e SPARKPOST_API_KEY=your-key -- npx -y @pcummings/sparkpost-mcp
Add --scope user to make it available in every project.
OpenAI Codex
codex mcp add sparkpost --env SPARKPOST_API_KEY=your-key -- npx -y @pcummings/sparkpost-mcp
Gemini CLI
gemini mcp add -e SPARKPOST_API_KEY=your-key sparkpost npx -y @pcummings/sparkpost-mcp
US region — append the base-URL env to any command above (-e for Claude/Gemini, --env for Codex):
SPARKPOST_API_BASE=https://api.sparkpost.com/api/v1
Any other MCP client (Claude Desktop, Cursor, Windsurf, VS Code, Cline, Zed …) — add this to the client's MCP config:
{
"mcpServers": {
"sparkpost": {
"command": "npx",
"args": ["-y", "@pcummings/sparkpost-mcp"],
"env": { "SPARKPOST_API_KEY": "your-key" }
}
}
}
For a local clone, use "args": ["tsx", "/absolute/path/to/index.ts"].
Run from source
npm install
export SPARKPOST_API_KEY=your-key # required; server exits if missing
export SPARKPOST_API_BASE=... # optional; defaults to EU. US: https://api.sparkpost.com/api/v1
npm start
Tools
| Tool | Description |
|---|---|
get_account |
Get SparkPost EU account info and usage |
list_templates |
List all email templates |
get_template |
Get a specific template by ID |
create_template |
Create a new email template |
update_template |
Update an existing template |
send_email |
Send a test/transactional email |
check_suppression |
Check if an address is suppressed |
list_sending_domains |
List all verified sending domains |
list_webhooks |
List all webhooks |
create_webhook |
Create a webhook (name, target URL, events) |
delete_webhook |
Delete a single webhook by id |
list_subaccounts |
List all subaccounts |
create_subaccount |
Create a subaccount with an API key (name, key_label, key_grants) |
search_message_events |
Search message events (one page; cursor + per_page for paging) |
get_deliverability_metrics |
Get deliverability metrics (from + metrics required; group_by selects a breakdown dimension) |
list_recipient_lists |
List all recipient lists |
get_recipient_list |
Get a recipient list by id (optionally with recipients) |
create_recipient_list |
Create a recipient list (recipients required; id/name auto-generated if omitted) |
add_suppression |
Add or update a suppression entry for one email |
remove_suppression |
Remove the suppression entry for one email |
Usage examples
Send an email:
Send a transactional email to alice@example.com from noreply@yourdomain.com with subject "Hello" and body "<p>Hi Alice</p>"
Create a template:
Create an email template with id "welcome", name "Welcome Email", subject "Welcome to {{company}}!", html "<p>Hi {{name}}</p>", from noreply@yourdomain.com
Search recent bounces:
Search message events for bounce events in the last 24 hours for recipient domain example.com
Dev
npm run typecheck # tsc --noEmit
npm test # node:test suite (44 tests)
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。