360 Secure SkillHub
360 安全技能中心

Automation Suggestion

该 Skill 在任务完成后判断是否满足自动化建议触发条件(7 天内重复至少 2 次、每周人工耗时至少 10 分钟或用户表达负面情绪),按固定话术至多每天提出一次建议;用户拒绝后 30 天内不再提示同类建议,并声称把建议与结果记入记忆用于学习。

szpili
szpili
热度 54.30
最近更新 2026-05-11 17:38:29
分类:自动化依赖:纯提示词应用场景:工作流 / 任务管理 / 推荐

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add automation-suggestion
Metadata
name
automation-suggestion
description
Suggest automations only when they make sense, based on objective triggers.
metadata
{"openclaw":{"emoji":"🤖","skillKey":"automation-suggestion"}}

Automation Suggestion Skill

Purpose

Suggest automations only when they make sense, based on objective triggers. Avoid nagging; only propose when clear value exists.

Triggers (any one qualifies)

  1. Repetition: Same task executed manually >= 2 times within 7 days
  2. Time saving: Estimated weekly manual effort >= 10 minutes
  3. User frustration: Detected via negative sentiment or explicit statement ("znowu to muszę robić", "to jest męczące")

Suggestion Format

After a task completes (when trigger detected):

[Automation idea]
I noticed you've done X 3 times this week (~15 min total).
Would you like me to automate it? I can set up a scheduled run or a quick command.
Reply: "yes" to configure, "no" to dismiss.

Rules

  • Never suggest before first manual completion
  • Max 1 suggestion per user per day
  • If user says "no", suppress similar suggestions for 30 days
  • Track suggestions and outcomes in memory for learning

Configuration (optional)

  • minRepetitions: 2
  • minWeeklyMinutes: 10
  • cooldownDaysAfterNo: 30

Integration

  • Hook: after task completion (via interaction-pipeline)
  • Reads task history from memory
  • Uses sentiment analysis on user messages (simple keyword-based)

Related Skills

  • interaction-pipeline (post-action hook)
  • satisfaction-learning (capture user response)