Agent Skillsaiming-lab/MetaClaw › agent-task-handoff

agent-task-handoff

GitHub

指导如何将子任务自包含地委托给子代理或跨会话移交,明确目标、上下文、约束、输入和成功标准,避免无上下文移交。

memory_data/skills/agent-task-handoff/SKILL.md aiming-lab/MetaClaw

Trigger Scenarios

将任务委派给子代理 跨会话移交工作 并行处理子任务

Install

npx skills add aiming-lab/MetaClaw --skill agent-task-handoff -g -y
More Options

Non-standard path

npx skills add https://github.com/aiming-lab/MetaClaw/tree/main/memory_data/skills/agent-task-handoff -g -y

Use without installing

npx skills use aiming-lab/MetaClaw@agent-task-handoff

指定 Agent (Claude Code)

npx skills add aiming-lab/MetaClaw --skill agent-task-handoff -a claude-code -g -y

安装 repo 全部 skill

npx skills add aiming-lab/MetaClaw --all -g -y

预览 repo 内 skill

npx skills add aiming-lab/MetaClaw --list

SKILL.md

Frontmatter
{
    "name": "agent-task-handoff",
    "category": "agentic",
    "description": "Use this skill when delegating a subtask to a sub-agent, spawning a parallel worker, or handing off work across sessions. Write a self-contained task description so the receiving agent needs no prior context."
}

Agent Task Handoff

When delegating work to a sub-agent or across sessions, make the handoff self-contained.

Handoff package must include:

  1. Goal: What the sub-agent must produce (one clear sentence).
  2. Context: The minimum background the agent needs — no more, no less.
  3. Constraints: Explicit rules (don't modify X, use library Y, output format Z).
  4. Inputs: File paths, API endpoints, or data the agent needs to start.
  5. Success criteria: How to verify the result is correct.

For parallel agents:

  • Clearly state which parts of the codebase/data each agent owns.
  • Define how outputs will be merged back.

Anti-pattern: Sending a vague "continue where we left off" with no context — sub-agents start from zero.

Version History

  • 922caf3 Current 2026-07-25 11:07

Same Skill Collection

memory_data/skills/async-communication-etiquette/SKILL.md
memory_data/skills/audience-aware-communication/SKILL.md
memory_data/skills/auth-and-authorization-patterns/SKILL.md
memory_data/skills/avoid-acting-on-assumptions/SKILL.md
memory_data/skills/avoid-hallucinating-specifics/SKILL.md
memory_data/skills/avoid-scope-creep/SKILL.md
memory_data/skills/clarify-ambiguous-requests/SKILL.md
memory_data/skills/codebase-navigation/SKILL.md
memory_data/skills/context-window-management/SKILL.md
memory_data/skills/data-validation-first/SKILL.md
memory_data/skills/debug-systematically/SKILL.md
memory_data/skills/do-not-retry-without-diagnosis/SKILL.md
memory_data/skills/git-workflow/SKILL.md
memory_data/skills/graceful-error-recovery/SKILL.md
memory_data/skills/idempotent-script-design/SKILL.md
memory_data/skills/input-validation-and-sanitization/SKILL.md
memory_data/skills/plan-before-multi-step-execution/SKILL.md
memory_data/skills/prioritize-by-impact/SKILL.md
memory_data/skills/professional-email-drafting/SKILL.md
memory_data/skills/robust-error-handling-in-scripts/SKILL.md
memory_data/skills/secrets-management/SKILL.md
memory_data/skills/secure-code-review/SKILL.md
memory_data/skills/source-evaluation/SKILL.md
memory_data/skills/sql-best-practices/SKILL.md
memory_data/skills/structured-logging-and-observability/SKILL.md
memory_data/skills/structured-progress-update/SKILL.md
memory_data/skills/structured-research-workflow/SKILL.md
memory_data/skills/structured-step-by-step-reasoning/SKILL.md
memory_data/skills/task-decomposition/SKILL.md
memory_data/skills/technical-writing-clarity/SKILL.md
memory_data/skills/test-before-ship/SKILL.md
memory_data/skills/tool-selection-strategy/SKILL.md
memory_data/skills/uncertainty-acknowledgment/SKILL.md
memory_data/skills/verify-before-irreversible-action/SKILL.md
memory_data/skills/visualization-selection/SKILL.md

Metadata

Files
0
Version
922caf3
Hash
8686b5fd
Indexed
2026-07-25 11:07

Главная - Вики-сайт
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-08-21 04:19
浙ICP备14020137号-1 $Гость$