← 返回 Skills 市场
2952
总下载
0
收藏
20
当前安装
1
版本数
在 OpenClaw 中安装
/install remind-me-2-1-0
功能描述
Set reminders using natural language. Automatically creates one-time cron jobs and logs to markdown.
安全使用建议
Do not install blindly. Key concerns: (1) The scripts call 'npx tsx' in /home/julian/clawdbot and will execute code from that repo (or fail) — ensure you control and inspect that project. (2) Reminders are delivered via Telegram to a hardcoded recipient (6636746252) — this will send reminder text externally and could leak sensitive info; confirm the recipient is intended. (3) The code requires undeclared tools ('npx', 'tsx', 'jq') and uses a hardcoded home path (/home/julian) making it user-specific. (4) There is an inconsistency: check-reminders.sh looks for '- [ ]' items, but create-* scripts log '- [scheduled]'/'- [recurring]', so automatic delivery may not work as written. Actions to consider before proceeding: ask the publisher to explain the Telegram target and provide required environment/dependency documentation; run the scripts in a safe isolated test user account after installing and inspect the local /home/julian/clawdbot code; or reject the skill until these inconsistencies and undeclared external deliveries are resolved.
功能分析
Type: OpenClaw Skill
Name: remind-me-2-1-0
Version: 1.0.0
The skill is classified as suspicious due to a potential argument injection vulnerability. The `create-reminder.sh` and `create-recurring.sh` scripts directly embed user-provided input (`MESSAGE`) into arguments for the `npx tsx src/index.ts cron add` command (e.g., `--name "Reminder: $MESSAGE"`). If the `cron add` command does not properly sanitize or quote these arguments, a malicious `MESSAGE` could lead to arbitrary command execution (RCE). The `SKILL.md` instructions for the agent also directly pass user input to these scripts, making the agent a vector for this vulnerability. There is no evidence of intentional malicious behavior like data exfiltration or persistence.
能力评估
Purpose & Capability
The description says it creates cron jobs and logs to markdown, which aligns with some files, but the scripts depend on a local project (cd /home/julian/clawdbot) and run 'npx tsx src/index.ts cron add' to create jobs and deliver messages via Telegram to a hardcoded recipient. The manifest declared only 'bash' and 'date' as required binaries; 'npx', 'tsx', and 'jq' are required by the scripts but not declared. Requiring a local Clawdbot repo and an external delivery channel is beyond the simple "cron + markdown" description and is not justified in the metadata.
Instruction Scope
SKILL.md instructs the agent to run the included scripts which read/write /home/julian/clawd/reminders.md and invoke the Clawdbot CLI (via npx) to deliver messages over Telegram to a fixed id (6636746252). The check-reminders.sh script expects unchecked items formatted as '- [ ] ...' and marks them '[x]', but create-reminder.sh and create-recurring.sh log entries as '- [scheduled]' and '- [recurring]', so the check script will not detect or mark those entries—an operational inconsistency. The scripts thus (a) modify user files in a hardcoded home path, (b) invoke a local JS tool with network/external-delivery side effects, and (c) will likely not work as intended due to inconsistent log formats.
Install Mechanism
There is no declared install spec (instruction-only), which is lower risk in isolation. However, the bundled scripts assume a node/npm environment ('npx', 'tsx') and 'jq' are available and that a local project exists at /home/julian/clawdbot. Those implicit dependencies increase runtime risk because arbitrary JavaScript code will be executed via npx from that repo (or will fail if absent).
Credentials
The skill declares no required environment variables or credentials but calls into a local Clawdbot command that delivers messages to Telegram (--channel telegram --to 6636746252). That implies the Clawdbot project must have credentials/config for delivery; those secrets are not declared. The hardcoded recipient ID means reminder content will be sent to a fixed external target (possible data leakage). No justification is provided for these credentials/targets in the SKILL.md.
Persistence & Privilege
always:false (normal) and the skill does not request system-wide privileges or modify other skills. It does, however, write to and edit a hardcoded file path (/home/julian/clawd/reminders.md) and assumes a particular user's directory layout; this limit of scope reduces broad privilege but makes the skill brittle and user-specific.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install remind-me-2-1-0 - 安装完成后，直接呼叫该 Skill 的名称或使用
/remind-me-2-1-0触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
- Initial release of Remind Me skill: set natural language reminders.
- Supports both one-time and recurring reminders via cron jobs.
- Logs all reminders to a markdown file for easy tracking.
- Understands both relative ("in 2 hours") and absolute ("tomorrow at 3pm") time formats.
- Provides manual commands to list, view, and remove reminders.
元数据
常见问题
Remind Me 2.1.0 是什么？
Set reminders using natural language. Automatically creates one-time cron jobs and logs to markdown. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 2952 次。
如何安装 Remind Me 2.1.0？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install remind-me-2-1-0」即可一键安装，无需额外配置。
Remind Me 2.1.0 是免费的吗？
是的，Remind Me 2.1.0 完全免费（开源免费），可自由下载、安装和使用。
Remind Me 2.1.0 支持哪些平台？
Remind Me 2.1.0 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Remind Me 2.1.0？
由 Lougazi（@loui1979）开发并维护，当前版本 v1.0.0。
推荐 Skills