Token导航 LogoToken导航

Agent Skill · 研究检索

remind-me-pro(提醒我亲)

remind-me-pro 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。

remind-me-pro 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为只读;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 2,159。

总安装 66,818周安装 2,703GitHub Stars 公开资料未说明下载 20,975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
0.1.0

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:remind-me-pro(提醒我亲)
来源仓库:https://github.com/youpele52/remind-me-pro
安装命令:
openclaw skills install remind-me-pro
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 remind-me-pro

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。

remind-me-pro 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词快速定位候选结果时使用。创建和管理频道内提醒与 cron 作业,限定作用范围。通过 clawhub 安装,命令为 openclaw skills install remind-me-pro。

建议确认权限范围、维护状态及是否触发联网操作。适用宿主包括 OpenClaw,接入前应确认版本、权限和运行环境要求。

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 remind-me-pro;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

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

remind-me-pro 怎么安装?

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

remind-me-pro 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。

remind-me-pro 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

remind-me-pro 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。

remind-me-pro 用的人多吗?

公开数据是累计安装 66,818 次,近一周安装 2,703 次,下载 20,975 次。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。

remind-me-pro 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。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 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

  • 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
remind-me-pro

提醒我亲 · 本页

66,818—只读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
remind-me
description
Create, list, and cancel reminders and cron jobs scoped to the channel they were requested from. Use when: user says 'remind me', 'set an alarm', 'schedule a cron', 'alert me when', 'every day at X do Y', 'cancel my reminder', 'list my reminders'. Auto-detects source channel and delivers back to it. Asks for clarification if schedule or intent is ambiguous before creating anything.
metadata
{"clawdbot":{"emoji":"⏰","requires":{"bins":["uv","openclaw"]}}}

Skill: Remind Me

When to use

  • User says "remind me to...", "set a reminder", "set an alarm"
  • User says "every Monday at 9 AM...", "every 30 minutes check..."
  • User says "in 20 minutes tell me...", "at 3 PM send me..."
  • User wants to schedule a recurring cron job via chat
  • User says "cancel my reminder", "delete that cron", "stop the alert"
  • User says "list my reminders", "what reminders do I have?"
  • User wants to know what cron jobs are active in this chat

When NOT to use

  • User wants to check a stock price right now (not scheduled) → use stock-price-checker-pro
  • User wants to run a one-time task immediately (no scheduling involved)

⚠️ CRITICAL: Always resolve these THREE things before creating a reminder

Before calling the script, you MUST have all three resolved:

FieldQuestion to answerExample
WHATWhat should happen / be said?"Check NVDA stock price"
WHEN / HOW OFTENOne-time or recurring? At what time/interval?"Every Monday at 9 AM"
WHEREWhich channel + chat ID to deliver to?Auto-detected from session

Missing field rules

  • WHAT is missing → Ask: "What would you like me to remind you about?"
  • WHEN is missing AND cannot be reasonably assumed → Ask: "How often, or at what time?"
  • WHEN is missing BUT can be reasonably assumed as once → Assume one-shot, but confirm: "Just once, right?"
  • WHERE is always auto-detected → Never ask the user for this. Read it from session context (see below).

Do NOT create the job until all three are confirmed.


Step 1 — Auto-detect channel and chat ID from session context

The source channel and chat ID are available in your session context. Extract them before doing anything else.

  • channel: the platform the message arrived on (e.g. telegram, discord)
  • chatId / to: the specific chat or user ID within that platform (e.g. <chatId>)

These two values are passed as --channel and --to to the script. Never ask the user for these. Never hardcode them. Always read from session context.


Step 2 — Parse the user's intent

From the user's natural language request, extract:

Schedule type

Map what the user said to one of three prefixed schedule strings:

What user saidSchedule string to pass
"every 30 minutes"every:30m
"every hour"every:1h
"every day at 9 AM"cron:0 9 * * *
"every Monday at 9 AM"cron:0 9 * * 1
"weekdays at 8 AM"cron:0 8 * * 1-5
"every Friday at 5 PM"cron:0 17 * * 5
"in 20 minutes"at:20m
"in 2 hours"at:2h
"at 3 PM today"at:<computed duration from now>
"once at 9:30 AM tomorrow"at:<computed duration from now>

One-shot vs recurring

  • at: → always one-shot (--once is auto-set by the script)
  • every: or cron: → recurring by default
  • If user says "just once" or "one time" with a cron: or every: → pass --once

Job name

Generate a short, descriptive name from the user's request.

  • "Remind me to check NVDA every Monday" → "NVDA Check - Monday 9AM"
  • "Alert me in 20 minutes" → "Alert - 20min"
  • "Grocery reminder at noon" → "Groceries - Noon"

Message

The message is what the agent will say or do when the job fires. Craft it clearly so the agent knows exactly what to do:

  • "Remind me to do groceries" → "Reminder: Time to do groceries! 🛒"
  • "Check NVDA every Monday at 9 AM" → "Check the current NVDA stock price and send me a summary."
  • "Send me a motivational quote every morning" → "Send me an inspiring motivational quote to start the day."

Step 3 — Clarification rules (ask before acting)

Ask when:

  1. No schedule at all: "Remind me to call John" → no time/frequency given

> Ask: "Sure! When would you like me to remind you — just once at a specific time, or on a recurring schedule?"

  1. Ambiguous frequency: "Remind me often" or "check regularly"

> Ask: "How often? Every hour, every day, or something else?"

  1. Conflicting signals: "Remind me every Monday but just once"

> Ask: "Just to confirm — should this be a one-time reminder or repeat every Monday?"

Do NOT ask when:

  • Everything is clear: "Remind me every Monday at 9 AM to check NVDA" → create immediately
  • One-shot is obvious from context: "Remind me in 20 minutes" → at:20m, once
  • User confirms after your clarification question → proceed immediately

Confirmation before creating (always):

Before calling the script, summarise what you're about to set up and get a quick confirmation:

"Got it! Here's what I'll set up: ⏰ Reminder: Check NVDA stock price 🔁 Schedule: Every Monday at 9 AM 📱 Delivered to: This chat Shall I go ahead?"

Only proceed after user confirms.


Commands

Create a reminder

uv run /root/.openclaw/workspace/skills/remind-me/src/main.py create \
  "<job name>" \
  "<what to say or do>" \
  "<every:duration | cron:expr | at:duration>" \
  "<channel>" \
  "<chatId>" \
  [once]

List reminders for this chat

uv run /root/.openclaw/workspace/skills/remind-me/src/main.py list \
  "<channel>" \
  "<chatId>"

Cancel a reminder by name

uv run /root/.openclaw/workspace/skills/remind-me/src/main.py cancel name "<job name>"

Cancel a reminder by ID

uv run /root/.openclaw/workspace/skills/remind-me/src/main.py cancel id "<job id>"

Examples

Example 1 — Clear request, no clarification needed

User: "Remind me every Monday at 9 AM to check NVDA"

  1. Detect: channel=telegram, to=<chatId>
  2. Parse: WHAT="Check NVDA stock price", WHEN=cron:0 9 * * 1, one-shot=false
  3. Confirm with user
  4. Run:
uv run /root/.openclaw/workspace/skills/remind-me/src/main.py create \
  "NVDA Check - Monday 9AM" \
  "Check the current NVDA stock price and send me a summary." \
  "cron:0 9 * * 1" \
  "telegram" \
  "<chatId>"

Example 2 — Missing frequency, ask first

User: "Remind me to go do groceries by 12 PM"

  1. Detect: channel=telegram, to=<chatId>
  2. Parse: WHAT="Do groceries", WHEN=12 PM but frequency unclear
  3. Ask: "Just to confirm — is this a one-time reminder for today at noon, or should I remind you every day at 12 PM?"
  4. User says: "Just today"
  5. Compute duration from now to today's noon → e.g. at:3h30m
  6. Confirm, then run:
uv run /root/.openclaw/workspace/skills/remind-me/src/main.py create \
  "Groceries - Noon" \
  "Reminder: Time to go do groceries! 🛒" \
  "at:3h30m" \
  "telegram" \
  "<chatId>" \
  once

Example 3 — List reminders

User: "What reminders do I have?"

uv run /root/.openclaw/workspace/skills/remind-me/src/main.py list \
  "telegram" \
  "<chatId>"

Format the output as a readable list, not raw JSON. Example response:

You have 2 active reminders in this chat: 1. NVDA Check - Monday 9AM 🔁 Every Monday at 9 AM _"Check the current NVDA stock price..."_ Next run: Mon 10 Mar 2026, 09:00 2. Groceries - Noon (one-time) _"Time to go do groceries!"_ Runs in: 3h 30m

Example 4 — Cancel a reminder

User: "Cancel my NVDA reminder"

  1. List jobs for this channel/chat first (internally)
  2. Find the matching job by name
  3. Confirm: "Cancel NVDA Check - Monday 9AM? This will stop all future runs."
  4. User confirms
  5. Run:
uv run /root/.openclaw/workspace/skills/remind-me/src/main.py cancel name "NVDA Check - Monday 9AM"

Example 5 — One-shot in-chat reminder

User: "In 30 seconds send me a love letter"

uv run /root/.openclaw/workspace/skills/remind-me/src/main.py create \
  "Love Letter - 30s" \
  "Write a beautiful, heartfelt love letter. Make it romantic and touching." \
  "at:30s" \
  "telegram" \
  "<chatId>"

Channel scoping rules

  • Jobs created from Group A are only visible when listing from Group A
  • Jobs created from Group B are only visible when listing from Group B
  • Jobs created from the main/private chat are visible only in that main chat
  • The script handles this automatically via the [remind-me:channel:chatId] tag embedded in each job's description
  • Never show a user jobs that belong to a different chat

Output formatting

After running any command, always format the result in plain conversational language — never dump raw JSON to the user.

On create success:

✅ Done! I've set up your reminder: ⏰ NVDA Check — Every Monday at 9 AM 📱 Delivered to this chat

On list (empty):

You have no active reminders in this chat.

On cancel success:

✅ Reminder "NVDA Check - Monday 9AM" has been cancelled.

On error:

❌ Something went wrong: <error message> Want me to try again?

Notes

  • uv run auto-installs dependencies from the inline script header — no pip or venv needed.
  • The script calls openclaw cron CLI internally — the gateway must be running.
  • Always use --channel "last" behaviour naturally: since --to is set to the originating chat ID, delivery is always back to the right place.
  • Do NOT use sessions_spawn, web search, or any other tool to create cron jobs — always go through this script.
  • Do NOT hardcode channel IDs. Always read from session context.
  • The at: schedule prefix does not support standard cron expressions — use cron: for those.

平台分布

OpenClaw

94.62%
按下载量换算19,847

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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