A lightweight real-time communication protocol for Kiro instances using shared JSON storage and polling.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install kiro-realtime
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 kiro-realtime 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 Kiro Realtime Chat?
Kiro Realtime is an essential addition to your collection of Openclaw Skills that enables seamless, near real-time interaction between multiple AI agent instances. By leveraging a shared JSON file as a communication bridge, different Kiro deployments—such as a cloud-based VPS and a local developer machine—can exchange messages, coordinate complex workflows, and share state updates.
This skill is particularly useful for developers building distributed agent systems who require a simple yet effective synchronization mechanism without the overhead of complex message brokers. By integrating this into your Openclaw Skills library, you provide your agents with the ability to communicate across different environments using a standardized messaging format.
Kiro Realtime Chat Use Cases
- Coordination of tasks across multiple Kiro instances in a distributed environment.
- Sending instant alerts or status updates between a VPS-hosted agent and a local machine.
- Triggering responses or secondary actions in one agent based on the output of another.
- Establishing a simple communication channel for multi-agent collaboration within the Openclaw Skills ecosystem.
How Kiro Realtime Chat Works
- A message is generated by a sender instance and appended to a shared JSON file located in the memory directory.
- The message includes unique identifiers such as ID, sender name, recipient name, timestamp, and a read status.
- Recipient instances perform periodic polling on the shared JSON file at defined intervals to check for new entries.
- When a new message addressed to the recipient is found, it is marked as read to prevent duplicate processing.
- The receiving instance processes the message and can optionally send a reply using the same lifecycle.
Kiro Realtime Chat Setup
To integrate this into your environment, ensure you have a shared filesystem or directory accessible by all Kiro instances. Initialize the storage directory and ensure your agent has write permissions.
# Create the memory directory if it doesn't exist
mkdir -p memory
# The skill will automatically create the JSON file on the first message,
# but you can initialize it manually if needed.
echo '{"messages": [], "lastCheck": ""}' > memory/kiro-realtime.json
Kiro Realtime Chat Data Schema & Taxonomy
The skill utilizes a structured JSON schema to maintain message integrity and tracking within the Openclaw Skills framework.
| Field | Type | Description |
|---|---|---|
messages |
Array | A list of message objects containing communication history. |
lastCheck |
ISO String | Timestamp of the last time an instance polled the file. |
id |
Integer | Unique identifier for each message. |
from / to |
String | Identifiers for the sender and recipient instances. |
read |
Boolean | Flag to indicate if the message has been processed. |
Kiro Realtime Chat Advanced Features
- Adjustable polling intervals ranging from 2 seconds for high-priority tasks to 60 seconds for low-bandwidth operations.
- Read-status tracking to ensure exactly-once processing of agent instructions.
- Timestamp-based filtering to optimize performance as the message history grows within Openclaw Skills.
- Extensible message objects that allow for custom metadata and complex task payloads.
SKILL.md
Loading
Related Openclaw Skills
Kiro Intercom
Kiro Intercom enables seamless communication and coordination between multiple agent instances through a shared markdown-based chat interface.
Agent Chat
A specialized messaging protocol that enables multiple Openclaw Skills instances to coordinate and exchange data through a shared JSON queue.
ClawMeter
A self-hosted dashboard for monitoring OpenClaw API spending and token usage with real-time budget alerts.
Improve Skill: The Meta-Skill
A meta-skill designed to evaluate and optimize other skills by identifying gaps in logic, codebase fit, and output quality.
Tweet Writer (twitterhdh)
A specialized AI agent skill designed to research, draft, and optimize viral tweets and threads based on real-time X algorithm data.
Blog Publisher
An automated workflow for transforming Obsidian notes into production-ready developer blog posts with AI-driven image processing and dual-language support.
Featured*
