Token导航 LogoToken导航TokenDH.com
效率只读clawhub未标认证来源可访问clear审计通过

rotating-single-target-cron旋转单目标 cron

Agent Skill

rotating-single-target-cron 用于补充效率相关能力,适合在 OpenClaw 中需要让 Agent 承接效率相关任务时使用。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。

总安装

5,936

周安装

255

GitHub Stars

公开资料未说明

下载量

2,081
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

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:rotating-single-target-cron(旋转单目标 cron)
来源仓库:https://github.com/sakullla/rotating-single-target-cron
安装命令:
openclaw skills install rotating-single-target-cron
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 rotating-single-target-cron

简介

创建轮换式聊天定时任务避免重复发送相同目标。

  • 每次运行时从预设 ID 列表中选取下一个接收方。
  • 保留最新调度记录确保任务连续性不受中断。rotating-single-target-cron 属于效率类 Skill,可作为该场景下的辅助能力补充。
  • 需预先配置目标 ID 列表与执行时间间隔参数。
  • 适用于群发通知或多用户轮询交互场景设计。

SKILL.md

name
rotating-single-target-cron
description
Create or update recurring chat cron jobs that choose exactly one @ID from a configured list on each run, avoid repeating the previous target, persist the last-picked target in a workspace state file, and post one single-target message to chat. Use when a user wants rotating daily/weekly mentions, story prompts, wake-up messages, teasing copy, or any other one-person-at-a-time scheduled content instead of group-wide output.

Rotating Single Target Cron

Overview

Use this skill to build or maintain cron jobs that rotate across a roster of @IDs. Each run should pick exactly one target, avoid the immediately previous target when alternatives exist, update a state file, and send only one final message.

Workflow

1. Confirm the shape of the job

Default assumptions for this pattern:

  • one recurring cron job
  • one message per run
  • one selected @ID per run
  • fixed roster unless the user asks to edit it
  • current chat as delivery target unless the user specifies another destination
  • last-picked target stored in a workspace file under memory/

Ask only for missing essentials:

  • roster of @IDs
  • schedule and timezone
  • message style or prompt brief
  • whether this is a new job or an update to an existing one

2. Prepare state

Create or maintain a plain text state file in the workspace. Recommended naming:

memory/<job-slug>-last-target.txt

State rules:

  • store exactly one line: @ID or none
  • if the file is missing, empty, or contains an ID no longer in the roster, treat it as none
  • when migrating an existing job, seed the file from the most recent known target if reliable; otherwise initialize it to none

3. Write a strict cron payload

The scheduled prompt should instruct the runtime to do this sequence:

  1. read the state file
  2. determine the previous target
  3. pick exactly one target from the configured roster
  4. avoid the previous target whenever another valid choice exists
  5. overwrite the state file with the newly chosen target
  6. output only the final chat message

Keep the payload strict about output shape. Unless the user says otherwise, include these rules:

  • mention the chosen @ID naturally near the beginning
  • do not mention the other configured IDs in the same message
  • keep the output public-chat safe
  • do not add explanations, labels, or meta commentary

4. Create or update the cron job

For a new job:

  • create the state file first
  • add the cron job with sessionTarget="isolated"
  • default delivery to the current chat
  • use a short descriptive name tied to the behavior

For an update:

  • patch the existing job in place
  • update schedule, roster, payload, and name as needed
  • do not create duplicates when the user is clearly editing an existing rotation

5. Test only on request

In active group chats, do not trigger a live run unless the user explicitly asks. If the user asks to test:

  • warn briefly that the test will post to the chat
  • run the cron job immediately
  • remember that the state file will advance after the test

6. Handle common edits safely

Common edits:

  • change the time
  • change the roster
  • swap the writing style or content brief
  • switch from multi-target message mode to rotating single-target mode
  • change word-count or tone constraints

When converting an older multi-target job into this pattern, initialize or repair the state file so the first rotated run starts cleanly.

Prompt skeleton

Use this as the base structure and replace placeholders with the user's style/content requirements.

Send exactly one message to the current chat every day at <HH:MM> (<TZ>).

Roster: <@id list>
State file: memory/<job-slug>-last-target.txt

You must follow these rules exactly:
1. Before generating anything, read the state file. If the file does not exist, is empty, or contains an ID not in the roster, treat the previous target as "none".
2. Randomly select exactly one person from the roster as the only target for this run.
3. Never pick the same target as last time when any other valid option exists.
4. After choosing the target, immediately overwrite the state file with that target's @ID.
5. Do not mention any other IDs from the roster in the final output.
6. Output only the final chat message. Do not add explanations, titles, labels, or commentary.
7. Mention the selected @ID naturally near the beginning of the message.
8. Follow these content requirements exactly:
<content brief and style constraints>

References

Read references/examples.md when you need ready-made prompt patterns for common variants such as heartbreak stories, wake-up copy, teasing posts, or fortune-style rotations.

Checklist

Before finishing, verify:

  • only one active cron job handles this rotation
  • schedule matches the user's request
  • roster matches the user's request exactly
  • state file uses a relative workspace path under memory/
  • prompt forbids repeating the immediately previous target
  • prompt forbids mentioning other configured IDs in the same output unless the user explicitly wants that
  • delivery target is correct
  • live testing happened only if the user requested it

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

74.09%
按下载量换算1,542

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。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills