handoff-skill

Agent 间任务交接的 Skill。当需要把任务交给另一个 Agent 时触发。触发词:交接、handoff、交给、让 xxx 来处理。

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "handoff-skill" with this command: npx skills add maxoreric/sop-engine/maxoreric-sop-engine-handoff-skill

任务交接

在 Agent 之间传递任务,参考邮件协议设计。

概念

邮件概念handoff 对应
From / To发送方 / 接收方 Agent
Subject任务主题
Date时间戳
Message-IDtask-id(唯一标识)
References上游任务(链式追溯)
Body任务描述
Attachmentsinputs/ 目录

流程

发送方

  1. 创建 task-id - 格式:{timestamp}-{random}
  2. 准备输入文件 - 复制到 inputs/
  3. 写 context.md - 填写 Header 和 Body
  4. 设置状态 - Status: pending
  5. 通知接收方 - 告知 task-id

接收方

  1. 读取 context.md - 理解任务
  2. 更新状态 - Status: in_progress
  3. 执行任务 - 根据 Body 描述执行
  4. 写入产出 - 放到 outputs/
  5. 更新状态 - Status: completed / failed

目录结构

agents/<agent-name>/handoff/<task-id>/
├── context.md       # 任务上下文(严格格式)
├── inputs/          # 输入文件
└── outputs/         # 输出文件

context.md 格式

# Handoff

## Header
- Message-ID: 20240101-abc123
- From: researcher-agent
- To: writer-agent
- Date: 2024-01-01T12:00:00Z
- Subject: 撰写调研报告
- References: 20231231-xyz789

## Body

请根据调研结果撰写一份报告,要求:
1. 结构清晰
2. 重点突出
3. 包含结论和建议

## Attachments

- inputs/research.md
- inputs/data.json

## Status

pending

Status 状态

  • pending - 等待接收方处理
  • in_progress - 接收方正在处理
  • completed - 已完成
  • failed - 失败

原则

  • context.md 格式严格,Hook 要读取
  • inputs/ 和 outputs/ 内容自由
  • References 形成任务链,支持追溯
  • 状态变更要及时更新

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

General

evaluate-skill

No summary provided by upstream source.

Repository SourceNeeds Review
General

reflect-skill

No summary provided by upstream source.

Repository SourceNeeds Review
General

clarify-skill

No summary provided by upstream source.

Repository SourceNeeds Review