handoff-chat

Append user messages to a shared handoff chat file for multi-agent coordination.

1|Updated Jan 6, 2026
One-click install
npx skills add https://github.com/Levezze/devkit --skill handoff-chat
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: handoff-chat
Source: https://github.com/Levezze/devkit/tree/main/skills/handoff-chat
Command: npx skills add https://github.com/Levezze/devkit --skill handoff-chat

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It solves the problem of keeping multiple AI agents synchronized on ongoing work without forcing the user to manually pass messages between repos and sessions.

Core Features & Use Cases

  • Cross-agent coordination via shared file: Writes conversation entries to a deterministic handoff chat file in /tmp so multiple agents can read context and continue discussion.
  • Stable identity and addressing: Records each agent participant under a stable handle derived from the repo and working directory (or a user-provided override), enabling targeted replies.
  • Topic-scoped slugging and session continuity: Auto-derives a human-decodable chat slug from the message subject and start timestamp, or continues an existing chat without re-deriving the slug.
  • Designed for lightweight back-and-forth: Use it for clarifying questions, status pings, and ongoing coordination, with the companion skill /handoff-reply for reading and responding.

Quick Start

Use the /handoff-chat skill to open a new or existing lightweight coordination thread by posting your message to the shared /tmp handoff chat file.

Frequently Asked Questions about handoff-chat

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

FAQPage Schema
How do I coordinate multiple AI agents across different repositories without manually passing messages?

Multi-agent coordination across repositories is handled by appending user messages to a shared handoff chat file, allowing agents to read context and continue discussion without manual user relay.

How does shared file communication work for ongoing agent back-and-forth?

Shared file communication works by appending messages to a deterministic chat file in /tmp, where multiple agents read ongoing context, clarifying questions, and status pings under stable participant handles.

What's the best way to maintain stable identity for agents participating in a cross-repo chat?

Stable identity for cross-repo chat participants is maintained by deriving a stable handle from the repository and working directory, or by applying a user-provided override for targeted replies.

Can I scope agent coordination threads to a specific topic and resume them later?

Topic-scoped coordination threads are created by auto-deriving a human-decodable chat slug from the message subject and start timestamp, allowing you to continue an existing chat without re-deriving the slug.

Does handoff chat require any external dependencies or components to start coordinating agents?

Agent coordination via handoff chat requires no external dependencies or components, relying solely on deterministic chat resolution, participant registry maintenance, and append-only message writing to a local file.

When should I not use a shared chat file for agent coordination?

Shared chat file coordination is designed for lightweight back-and-forth, clarifying questions, and status pings, making it unsuitable for heavy data transfers or complex state management outside of ongoing text-based context sharing.