chat-over-session

Coordinate multi-agent chat sessions through a shared append-only Markdown file.

1|Updated Feb 23, 2026

Search Tags:#collaboration #communication #multi-agent #real-time #chat #file sharing

Ready to run

npx skills add https://github.com/kang-sw/devenv --skill chat-over-session

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: chat-over-session
Source: https://github.com/kang-sw/devenv/tree/main/claude-skills/chat-over-session
Command: npx skills add https://github.com/kang-sw/devenv --skill chat-over-session

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill enables multiple AI agents to communicate and collaborate on a shared topic by discussing it through a common file, facilitating complex problem-solving and idea generation.

Core Features & Use Cases

Multi-Agent Communication: Agents can join or create chat sessions to discuss topics.
Shared File Collaboration: All communication happens within a designated Markdown file, acting as a persistent log.
Use Case: Imagine a team of agents brainstorming marketing strategies. One agent can initiate a chat, others can join, and they can collectively refine ideas, assign tasks, and document decisions within the shared chat file.

Quick Start

Use the chat-over-session skill to join the discussion with the agent name "Researcher".

Frequently Asked Questions about chat-over-session

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

FAQPage Schema

How do I enable real-time chat between multiple AI agents?▼

Multi-agent chat is facilitated through a shared, append-only Markdown file that logs all messages. A custom bash script monitors the file in real-time to deliver updates, allowing agents to join or create sessions and collaborate on shared topics.

What is the best way to track conversation progress in a multi-agent chat session?▼

Conversation progress in a multi-agent chat session is tracked using a robust 'fence' mechanism. This ensures no messages are lost during real-time communication, maintaining a reliable and persistent log of all agent interactions.

Can I use a Markdown file for real-time multi-agent collaboration?▼

Yes, a designated Markdown file acts as the persistent log for real-time multi-agent collaboration. Agents append their messages to this shared file, which a bash script monitors to ensure immediate message delivery and synchronized communication.

How do multiple AI agents join the same communication session?▼

Agents join the same communication session by accessing a shared chat file. One agent initiates the session, and others join in to collectively refine ideas, assign tasks, and document decisions within the append-only Markdown log.

Why use an append-only file for multi-agent communication?▼

An append-only Markdown file is used for multi-agent communication to provide a persistent, shared log. This prevents data loss, allowing multiple agents to safely append messages, track progress with a fence mechanism, and collaborate effectively.

Does multi-agent chat over sessions require external dependencies?▼

No, this multi-agent chat approach does not require external dependencies. It relies on a custom bash script for real-time file monitoring and a shared Markdown file, ensuring a self-contained environment for agent collaboration.