How does it handle complex tasks with many dependencies?

During the 'Create' mode, the skill performs a pre-creation analysis that identifies task separations and uses dependency blocking instructions to ensure the next session starts with a clear priority list.

Can I choose where the handover report is saved?

Yes, you can use the 'file' argument to write exclusively to a document, 'chat' to display only in the terminal, or use no arguments to perform both actions simultaneously.

How does the Handover skill detect whether to resume or save work?

The skill automatically analyzes the current conversation state: a minimal or empty session history triggers 'Process' mode to resume work, while an active session with progress triggers 'Create' mode to document the handover.

What is the role of Haiku subagents in this skill?

Haiku subagents are utilized during 'Process' mode to quickly and efficiently scan through previous session handovers to locate the specific point where work needs to be resumed.

Session Handover Manager

Name: Session Handover Manager
Author: Mharbulous

byMharbulous

生産性とワークフロー

Maintains project continuity by automatically resuming previous tasks or documenting current progress for future coding sessions.

概要

The Handover skill provides a structured workflow for maintaining context between Claude Code sessions, solving the problem of context loss during multi-stage development cycles. It intelligently detects whether to resume a task by using Haiku subagents to cascade through history and find the last incomplete milestone, or to document current progress by analyzing conversation states and task dependencies. By enforcing a standardized 'Create' and 'Process' cycle, it ensures that complex software engineering tasks can be paused and resumed without losing critical implementation details or architectural decisions.

主な機能

Human-decision workflow integration for subagent recommendations
Pre-creation analysis for task separation and dependency blocking
Recursive cascading through handovers via Haiku subagents to find incomplete work
0 GitHub stars
Automatic mode detection based on conversation state and history
Flexible output routing to chat, local files, or both

ユースケース

Managing state for long-running feature development that spans multiple AI interactions
Resuming complex refactoring tasks after a session break without losing architectural context
Generating standardized progress reports and next steps for asynchronous team handoffs

概要

主な機能

Human-decision workflow integration for subagent recommendations
Pre-creation analysis for task separation and dependency blocking
Recursive cascading through handovers via Haiku subagents to find incomplete work
0 GitHub stars
Automatic mode detection based on conversation state and history
Flexible output routing to chat, local files, or both

ユースケース

Managing state for long-running feature development that spans multiple AI interactions
Resuming complex refactoring tasks after a session break without losing architectural context
Generating standardized progress reports and next steps for asynchronous team handoffs