skills.rest
zackham
zackhamCOMMUNITY
Explore all skills by this author →
GitHub

handoff

Parse JSONL transcripts and generate structured handoff documents for session resumption.

Updated Feb 25, 2026
Search Tags:#session management#handoff#context preservation#state saving#crash recovery#ai continuity
One-click install
npx skills add https://github.com/zackham/agent-skills --skill handoff-zackham
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: handoff
Source: https://github.com/zackham/agent-skills/tree/main/handoff
Command: npx skills add https://github.com/zackham/agent-skills --skill handoff-zackham

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill prevents the loss of work due to session timeouts, context window limits, or accidental closures, ensuring continuity and preventing data loss in AI coding sessions.

Core Features & Use Cases

  • Session Handoff: Voluntarily save the current state of an AI session into a structured document.
  • Crash Recovery: Automatically reconstruct session context from transcripts after a crash or unexpected exit.
  • Contextual Continuity: Enables seamless resumption of work across different sessions, preserving critical decisions and progress.
  • Use Case: If your AI coding session crashes mid-task, you can use /recover to load the last saved state and continue exactly where you left off, without losing any code or context.

Quick Start

Use the handoff skill to save the current session state by typing /handoff.

Frequently Asked Questions about handoff

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I save my AI coding session state before a context window limit or timeout?▼

To save your AI session state, type `/handoff` to voluntarily synthesize the current context into a structured document, ensuring seamless resumption and preventing data loss across session boundaries.

Can I recover an AI session after an unexpected crash or closure?▼

Yes, you can recover an AI session after a crash by using `/recover` to automatically reconstruct session context from JSONL transcripts, extracting key information to load the last saved state and continue mid-task.

How does session context preservation work for AI coding environments?▼

Session context preservation works by parsing JSONL transcripts and extracting key information through a multi-phase LLM process, which generates a structured handoff document that captures critical work, decisions, and progress for continuity.

Does the handoff process support filtering for specific focus areas?▼

Yes, the handoff process supports optional focus filtering, allowing you to preserve only critical work and decisions across session boundaries rather than reconstructing the entire raw transcript.

What is the best way to maintain AI continuity across multiple coding sessions?▼

The best way to maintain AI continuity is to use a structured handoff document that synthesizes and reconstructs session context, allowing you to voluntarily save the current state and resume exactly where you left off.

Do I need any dependencies to manage crash recovery for AI sessions?▼

No dependencies are required to manage crash recovery for AI sessions, as the Skill independently parses JSONL transcripts and extracts key information to generate the structured handoff document.

Related Skills

Explore More Software Engineering Skills→
aresbitaresbit
46

session-logs

Search and analyze historical session JSONL files using jq and rg.

Community
Intermediate
the-inconvenience-storethe-inconvenience-store
1

session-handoff

Create a session handoff document with git status and branch information.

Community
Advanced
kivo360kivo360
1

moai-alfred-session-state

Manage Alfred session state across workflows with session_state.json persistence.

Community
Advanced
AJBcodingAJBcoding
4

moai-cc-memory

Manage session memory and persist context across Claude Code sessions.

Community
Advanced