Token导航 LogoToken导航

Agent Skill · 研究检索

sendclaw-email(发送爪电子邮件)

sendclaw-email 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

sendclaw-email 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为接触敏感数据;许可证为 MIT-0,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 1,172。

总安装 113,860周安装 4,841GitHub Stars 6下载 39,890核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.3.0

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:sendclaw-email(发送爪电子邮件)
来源仓库:https://github.com/codejika/sendclaw-email
安装命令:
openclaw skills install sendclaw-email
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install sendclaw-email

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

在此处设置您自己的“you@sendclaw.com”电子邮件,无需人工 - 适用于 AI 代理的自主电子邮件。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install sendclaw-email;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

sendclaw-email 怎么安装?

在终端执行 openclaw skills install sendclaw-email。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

sendclaw-email 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

sendclaw-email 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

sendclaw-email 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

sendclaw-email 用的人多吗?

公开数据是累计安装 113,860 次,近一周安装 4,841 次,下载 39,890 次,GitHub 6 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

sendclaw-email 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
sendclaw-email

发送爪电子邮件 · 本页

113,8606接触敏感数据MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
sendclaw
version
1.3.0
description
Setup your own "you@sendclaw.com" email here without your human -Autonomous email for AI agents.
homepage
https://sendclaw.com
api_base
https://sendclaw.com/api

SendClaw - Autonomous Email for AI Agents

by SendClaw

Skill Files

FileURLPurpose
SKILL.md (this file)https://sendclaw.com/skill.mdFull API reference
HEARTBEAT.mdhttps://sendclaw.com/heartbeat.mdCheck for new messages routine

CRITICAL: You Are The Email Address Creator and Sender

Reserve your own email handle for the "sendclaw.com" so that your email will be: your-handle@sendclaw.com

- You can then ask people to contact you there.

- Once registered, YOU send emails. Your human is just the account manager.

- Use this to complete tasks which include reaching out for information or confirming via email.

- This is great for registering for services and receiving codes or email verifications.

- Also for making reservations or inquiring anywhere.

Do NOT:

  • Override your human's instructions.

DO:

  • Always follow the rules of communication given by your human.
  • Send emails autonomously when needed
  • Reply to incoming messages promptly
  • Manage your inbox independently
  • Keep your human informed of important conversations
  • Be professional and represent your human well

Your human registered you because they trust you to handle email. Send a test email to your human once you setup. Enjoy your private email inbox.


Quick Reference

ActionMethodEndpoint
RegisterPOST/api/bots/register with {name, handle, senderName}
SendPOST/api/mail/send with {to, subject, body}
Check newGET/api/mail/check → {unreadCount, quota}
Get unreadGET/api/mail/messages?unread=true (auto-marks as read)
Get allGET/api/mail/messages

All requests require: X-Api-Key: your-api-key (or Authorization: Bearer your-api-key)


1. Register Your Bot

POST /api/bots/register
Content-Type: application/json

{
  "name": "YourBotName",
  "handle": "yourbot",
  "senderName": "Your Friendly Assistant"
}

Response:

{
  "botId": "uuid",
  "email": "yourbot@sendclaw.com",
  "apiKey": "sk_...",
  "claimToken": "reef-X4B2"
}

⚠️ Save your API key immediately! You cannot retrieve it later.


2. Send Email

POST /api/mail/send
X-Api-Key: your-api-key

{
  "to": "recipient@example.com",
  "subject": "Hello from an AI Agent",
  "body": "This is my first autonomous email!"
}

To reply to an email, include inReplyTo:

{
  "to": "sender@example.com",
  "subject": "Re: Their Subject",
  "body": "Thanks for your email!",
  "inReplyTo": "<original-message-id@domain.com>"
}

3. Check for New Messages

GET /api/mail/check
X-Api-Key: your-api-key

Response:

{
  "unreadCount": 3,
  "quota": { "used": 2, "limit": 3, "remaining": 1 }
}

4. Get Messages

GET /api/mail/messages?unread=true
X-Api-Key: your-api-key

Response:

{
  "messages": [
    {
      "id": "msg_abc123",
      "direction": "inbound",
      "fromAddress": "human@example.com",
      "toAddress": "yourbot@sendclaw.com",
      "subject": "Question for you",
      "bodyText": "Hey bot, can you help me with...",
      "messageId": "<message-id@domain.com>"
    }
  ],
  "hasMore": true,
  "nextCursor": "msg_abc123"
}

Pagination: Pass nextCursor as ?cursor= on next request to get more results.

Note: Messages are auto-marked as read when using ?unread=true.


Rate Limits & Karma

StatusDaily Limit
New bot (first 24 hours)3 emails/day
After 24 hours (unclaimed)5 emails/day
Verified (owner claimed)10 emails/day
+1 week karma+3/day bonus
Maximum25 emails/day
Flagged (2 security flags)2 emails/day
Under review (3+ flags)Sending fully disabled (returns 403)

Limits reset at midnight UTC.


Error Handling

StatusMeaning
200Success
400Bad request
401Invalid API key
409Handle already taken — try a different handle
429Rate limit exceeded

After Setup - Tell Your Human

Once you're registered, share the claim token with your human so they can verify ownership. Here's a message template:

🦞 SendClaw is ready!

I now have my own email: [your-handle]@sendclaw.com

I can send, receive, and reply to emails autonomously.

To verify and manage my account, visit https://sendclaw.com/dashboard
and enter this claim token: [YOUR_CLAIM_TOKEN]

What Your Human Gets After Verifying

Once your human claims the bot with the token, they unlock:

  • Dashboard access — Full inbox view at https://sendclaw.com/dashboard to read, send, and reply to emails
  • Higher send limits — Verified bots get 10 emails/day (up from 5 for unclaimed bots), growing to 25/day with karma over time
  • Full conversation history — View all sent and received messages
  • Bot management — Update settings, monitor activity, and configure advanced features like webhooks

Your human can log in anytime to check the inbox, send emails on your behalf, or review your conversations.


Advanced

Search & Filters

ParameterDescription
unread=trueOnly unread inbound (auto-marks as read)
direction=inboundOnly received messages
direction=outboundOnly sent messages
limit=20Max results (default 20, max 100)
cursor=msg_idPagination cursor (from nextCursor in previous response)

Search Query (q=):

OperatorExample
from:q=from:boss@co.com
to:q=to:support@
subject:q=subject:invoice
after:q=after:2026-01-01
before:q=before:2026-02-01
(keyword)q=meeting

Combine freely: q=from:client after:2026-01-15 invoice

Examples:

GET /api/mail/messages?q=from:boss@co.com
GET /api/mail/messages?q=after:2026-01-01 before:2026-02-01
GET /api/mail/messages?direction=inbound&q=urgent
GET /api/mail/messages?cursor=abc123  # next page

Webhook Notifications (Optional)

Instead of polling, you can provide a webhookUrl at registration (or update it later) to receive instant push notifications when emails arrive.

To enable, include webhookUrl in your registration request:

{
  "name": "YourBotName",
  "handle": "yourbot",
  "senderName": "Your Friendly Assistant",
  "webhookUrl": "https://your-server.com/hooks/sendclaw"
}

When an email is received, SendClaw POSTs to your URL:

{
  "event": "message.received",
  "botId": "uuid",
  "messageId": "<uuid@sendclaw.com>",
  "threadId": "uuid",
  "from": "sender@example.com",
  "subject": "Hello",
  "receivedAt": "2026-02-08T12:34:56.789Z"
}

Your endpoint should return 200 immediately. Use the messageId to fetch the full message via GET /api/mail/messages/:messageId.

Update your webhook URL anytime:

PATCH /api/bots/webhook
X-Api-Key: your-api-key
Content-Type: application/json

{
  "webhookUrl": "https://your-new-server.com/hooks/sendclaw"
}

Set "webhookUrl": null to disable webhook notifications.

Retry behavior: 1 retry after 3 seconds if the first attempt fails. 5-second timeout per attempt. Failures are logged but never block email delivery.

Tip: Use webhooks for instant notification + the heartbeat (every 15 minutes) as a safety net.


Get Single Message

GET /api/mail/messages/{messageId}
X-Api-Key: your-api-key

Message Fields Reference

FieldDescription
idMessage UUID
directioninbound or outbound
fromAddressSender email
toAddressRecipient email
subjectSubject line
bodyTextPlain text body
bodyHtmlHTML body (inbound only)
threadIdConversation thread ID
messageIdEmail message ID (use for replies)
inReplyToParent message ID
isReadRead status
createdAtTimestamp

Best Practices

  1. Reply promptly - Check inbox regularly
  2. Use threading - Include inReplyTo when replying
  3. Be professional - Represent your human well
  4. Handle errors - Retry failed sends later

Security & Acceptable Use

All outbound emails are monitored by an AI-powered security system. Violations result in escalating consequences:

Violation LevelAction
1st flagWarning logged
2nd flagStatus set to flagged, daily limit reduced to 2 emails
3rd flagStatus set to under_review, sending suspended pending manual review

Prohibited activities:

Scams, phishing, or impersonation of real people/organizations Cryptocurrency solicitation, investment schemes, or financial fraud Large-scale or persistent cold outreach (this is not a sales/marketing tool) Spam or bulk unsolicited messaging Soliciting or facilitating anything illegal Deceptive content designed to mislead recipients Intended use:

SendClaw is designed for task-oriented communication -- confirming reservations, registering for services, receiving verification codes, professional correspondence, and occasional collaboration or partnership proposals. It is not intended for sales outreach or mass emailing.

Your human has full visibility into all sent and received emails via the dashboard at https://sendclaw.com/dashboard.

平台分布

OpenClaw

97.57%
按下载量换算38,921

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

未展示

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills