mycelium

Centralize handover of work context between agents, accounts, and devices.

Updated May 15, 2026
One-click install
npx skills add https://github.com/heyu-ai/yibi-stack --skill mycelium-heyu-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mycelium
Source: https://github.com/heyu-ai/yibi-stack/tree/main/plugins/growth/skills/mycelium
Command: npx skills add https://github.com/heyu-ai/yibi-stack --skill mycelium-heyu-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill provides a unified handover and insight system for multi-agent collaboration, ensuring seamless information transfer and knowledge retention across different agents, accounts, and devices.

Core Features & Use Cases

  • Unified Handover: Centralized handover of work context between agents, accounts, and devices.
  • Insight Collection: Automatically collects insights and away summaries from Claude Code sessions.
  • Data Migrations: Facilitates migration of old data from previous systems.
  • Account Detection: Provides multi-layer fallback for account detection across different environments.
  • Use Case: For a team using Claude Code across multiple devices and accounts, this Skill ensures that work context is preserved and insights are easily accessible.

Quick Start

Initialize the mycelium skill by running the command: uv run --directory "$SKILL_REPO" python -m tasks.mycelium init --device-id {{device_id}} --default-account {{default_account}}.

Frequently Asked Questions about mycelium

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

FAQPage Schema
How do I retain work context when switching between multiple Claude Code accounts and devices?

Multi-agent collaboration context retention is achieved by transferring work handovers and away summaries across different agents, accounts, and devices, ensuring seamless information transfer and knowledge preservation without manual duplication.

How does automated insight collection work for Claude Code sessions?

Automated insight collection works by extracting away summaries and session data from Claude Code environments, centralizing the knowledge into a unified system so that subsequent agents can access previous context immediately.

Can I migrate old data from previous systems into a new multi-agent collaboration workflow?

Yes, data migration from previous systems is supported natively, allowing you to transfer historical work context and insights into the unified handover system to maintain continuity across updated agent environments.

How do I initialize a unified handover system for multi-agent collaboration?

To initialize the handover system, run the provided Python init module with your specific device ID and default account parameters, establishing the baseline environment for seamless multi-device context synchronization.

Does multi-layer fallback account detection work across different environments?

Account detection provides multi-layer fallback across different environments, accurately identifying user accounts even when primary configuration methods fail, ensuring uninterrupted context handover and insight collection.