session-handoff

Package session observations and MEMORY.md state into a structured Handoff Block.

Updated Jun 2, 2026
One-click install
npx skills add https://github.com/chicago-joe/manymems --skill session-handoff-chicago-joe
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: session-handoff
Source: https://github.com/chicago-joe/manymems/tree/main/plugin/skills/session-handoff
Command: npx skills add https://github.com/chicago-joe/manymems --skill session-handoff-chicago-joe

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Package the current session context into a structured handoff block so another agent can resume work without starting from scratch.

Core Features & Use Cases

  • Gather Session Observations: Retrieve all tool calls, file edits, and decisions from the current session.
  • Gather MEMORY.md State: Include the latest persistent project context for continuity.
  • Output Handoff Block: Produce a structured block that includes Current Task, In-Flight Files, Completed This Session, Next Steps, Open Questions, and Relevant Memory.

Quick Start

Start by collecting the current session observations and MEMORY.md, then output a Handoff Block in the specified format.

Frequently Asked Questions about session-handoff

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

FAQPage Schema
How do I hand off an AI agent session to another agent without losing context?

Multi-agent session handoff works by collecting tool calls, file edits, and decisions from the current session, then formatting them into a structured block for another agent to resume work.

What is the best way to save session state when context limits are reached?

The best way to save session state at context limits is to generate a structured handoff block that captures current tasks, completed actions, next steps, and persistent MEMORY.md state for later resumption.

What is included in a structured session handoff block?

A structured session handoff block includes the Current Task, In-Flight Files, Completed This Session, Next Steps, Open Questions, and Relevant Memory, ensuring another agent can resume work immediately.

When should I package context for a new agent session?

You should package context when a user asks to hand off, continue in a new session, save session state, or when context limits are reached during an active multi-agent workflow.