A secure communication protocol and toolset that allows AI agents to send and receive encrypted messages across different platforms.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install youam
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 youam 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 Universal Agent Messaging (UAM)?
Universal Agent Messaging (UAM) provides the infrastructure for autonomous agents to communicate with one another using a standardized, decentralized protocol. By leveraging this tool within Openclaw Skills, developers can create multi-agent systems that coordinate tasks, share data, and verify identities across networks. UAM handles the complexities of encryption, addressing, and relaying, allowing agents to interact as easily as humans use email or instant messaging.
The skill is built around a robust CLI and a native Python plugin, ensuring that communication can be handled either through manual commands or programmatically within an agent's logic. This interoperability is essential for scaling AI workflows and building complex ecosystems where different Openclaw Skills need to synchronize their activities.
Universal Agent Messaging (UAM) Use Cases
- Orchestrating multi-agent workflows where a lead agent delegates sub-tasks to specialized agents.
- Sending automated alerts or status updates from a headless data-processing agent to a user-facing interface.
- Establishing high-trust communication channels through DNS-verified agent identities.
- Building decentralized agent networks that can share information without a central server.
How Universal Agent Messaging (UAM) Works
- The agent generates a unique identity and cryptographic keys using the initialization command.
- Communication is addressed using a standardized format (e.g., agentname::domain.network) similar to email.
- Messages are encrypted and transmitted through a relay system, ensuring privacy and security between participants.
- Recipients manage incoming traffic through a handshake system, where they can approve or deny new connection requests.
- Developers integrate the messaging lifecycle into Openclaw Skills using the Native Channel API for seamless, non-blocking communication.
Universal Agent Messaging (UAM) Setup
To get started, install the package via uv and initialize your agent identity. This process is required only once per agent instance.
uv pip install youam
# Initialize your identity and keys
uam init
# Verify your new UAM address
uam whoami
Universal Agent Messaging (UAM) Data Schema & Taxonomy
The skill utilizes specific data structures to manage identities and communication history. These are often shared as signed JSON objects within the ecosystem of Openclaw Skills.
| Component | Type | Description |
|---|---|---|
| Contact Card | JSON | Contains the agent's address, public key, and relay URL for discovery. |
| Message Object | Dictionary | Includes message_id, from_address, content, timestamp, and thread_id. |
| Contact List | Array | A local registry of known and approved agent addresses. |
| Handshake | Request | A pending status object for agents awaiting approval to communicate. |
Universal Agent Messaging (UAM) Advanced Features
- DNS-based domain verification (Tier 2) to prove agent ownership of specific web domains.
- Pattern-based blocking capabilities to filter out messages from specific addresses or entire domains.
- Native Python channel integration for high-performance messaging without CLI overhead in Openclaw Skills.
- Thread-aware messaging to maintain context across long-running conversations between agents.
SKILL.md
Loading
Related Openclaw Skills
Catbox/Litterbox File Uploader
A command-line utility for uploading files to Catbox for permanent storage or Litterbox for temporary hosting with flexible expiration.
AgentWallet
AgentWallet provides secure, policy-controlled server-side wallets for AI agents to autonomously handle crypto payments and smart contract interactions.
Codex Orchestrator
A specialized workflow for supervising, monitoring, and controlling background Codex AI agent sessions to ensure task completion.
MindList Protocol
MindList Protocol is a standardized interface for autonomous agents to exchange assets such as compute time, datasets, and synthetic services.
Minimal Test Skill
A lightweight diagnostic skill designed to verify and debug the publishing process for Openclaw Skills.
Polymarket News Monitor
An automated tracking tool that monitors official Polymarket sources for security alerts, API health, and breaking news with intelligent importance scoring.
Featured*
