A zero-dependency Python messaging bus that enables users and AI agents to communicate through shared file directories.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install chat-bus
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install chat-bus using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Chat Bus?
Chat Bus is a lightweight, file-system-based communication protocol designed to facilitate interaction between different users and AI agents. By utilizing a shared directory—such as a NAS, OneDrive, Syncthing, or an SMB mount—it creates a robust messaging backbone without the need for a centralized server. As a featured tool within the Openclaw Skills ecosystem, Chat Bus allows for private messaging, group rooms, and broadcasts, making it an ideal choice for developers seeking a decentralized communication layer.
The skill is built entirely using the Python standard library, ensuring zero external dependencies and native cross-platform compatibility across Windows, macOS, and Linux. It transforms every message into a persistent JSON file, providing natural history tracking and easy integration for any agent capable of file system operations.
Chat Bus Use Cases
- Establishing a communication channel between multiple autonomous AI agents in a local network.
- Facilitating human-in-the-loop interactions with agents via shared cloud storage like Dropbox or iCloud.
- Creating serverless chat rooms for small teams or IoT devices using existing file synchronization tools.
- Auditing and archiving agent conversations through standard file system backups.
How Chat Bus Works
- Users and agents must first perform a registration step to create a unique identity profile within the shared directory.
- Communication is initiated by writing a JSON-formatted message file to a specific destination folder (inbox or room).
- The receiving agent or user monitors the shared directory for new files to ingest incoming data.
- Message state is managed through file naming conventions, where read status can be tracked by appending specific extensions.
- Group logic is handled by room configurations that define membership and topic metadata for collective discussions.
Chat Bus Setup
To integrate this into your workflow, ensure all participating agents have read/write access to a common folder. Use the following commands to initialize the Openclaw Skills environment:
# Register an agent or user
python register.py '{"user":"agent_one","display_name":"Coding Assistant"}'
# Send a private message
python send.py '{"user":"agent_one","to":"human_user","content":"Task completed successfully."}'
# Join a shared room
python rooms.py '{"action":"join","user":"agent_one","room":"dev_squad"}'
Chat Bus Data Schema & Taxonomy
Chat Bus organizes data within a hidden .chat-bus/ directory with the following structure:
| Path | Content Description |
|---|---|
users/ |
Contains [username].json files storing user metadata and display names. |
inbox/[user]/ |
Stores private messages. Files ending in .read represent acknowledged messages. |
rooms/[room]/ |
Contains group messages and a _room.json file for membership and topic management. |
history/ |
Aggregated view of historical JSON message logs for long-term storage. |
Chat Bus Advanced Features
- Decentralized multi-agent support without a central API or message broker.
- Cross-platform synchronization leveraging existing enterprise file-sharing infrastructure.
- Human-readable JSON message format for easy debugging and manual intervention.
- Native persistence ensures no messages are lost during agent downtime or restarts.
- Extensible room management allowing for dynamic topic updates and member lists.
SKILL.md
Loading
Related Openclaw Skills
PRD-Writer
A high-performance AI skill designed to transform raw ideas or fragments into comprehensive, audit-ready Product Requirement Documents.
RingCentral for AI Agents
A powerful communication skill that enables AI agents to manage RingCentral calls, messages, contacts, and team collaboration via the Membrane framework.
Dida365 Integration
A powerful API integration for Dida365 that enables automated task and project management through official OAuth authentication.
HNA 666Plus Flight Checker
An automated browser-based tool to scan Hainan Airlines 666Plus benefit availability for round-trip flights departing from Beijing.
Skill Analyst
A professional evaluation tool for analyzing, comparing, and vetting AI agent skills before installation or publication.
Custodian
Custodian is an autonomous maintenance agent that detects, classifies, and repairs Openclaw operational failures during quiet hours.
Featured*
