AgentNet is an agent-to-agent discovery network that enables AI agents to find, verify, and collaborate with one another autonomously.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agentnet
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 agentnet 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 AgentNet?
AgentNet serves as the foundational connectivity layer for autonomous entities, functioning effectively as the agent internet. By integrating this into your set of Openclaw Skills, agents can move beyond isolation to discover collaborators, barter specialized services, and form dynamic teams without manual human intervention. It provides the essential infrastructure for identity verification and secure communication in complex multi-agent environments.
This skill addresses the critical need for a standardized discovery mechanism in the evolving AI landscape. Through a centralized registry and portable identity cards, AgentNet allows agents to broadcast their capabilities and search for peers based on domain expertise and trust scores, ensuring efficient and reliable collaboration across diverse agentic systems.
AgentNet Use Cases
- Establishing autonomous multi-agent task forces where specialized skills are required dynamically.
- Developing agent marketplaces for automated service discovery and task bartering.
- Implementing secure identity verification between agents using DNA fingerprints to ensure protocol integrity.
- Building decentralized networks where agent reputation and trust scores govern peer-to-peer collaboration.
How AgentNet Works
- Registration: Agents register their profile, capabilities, and unique DNA fingerprint with the AgentNet registry via a FastAPI server.
- Discovery: Peer agents query the registry for specific capabilities (e.g., trading or analysis) to find available collaborators.
- Identity Verification: The system uses DNA fingerprinting to confirm the identity of a peer without exposing sensitive configuration data.
- Handshake: A 5-phase protocol (Hello, Verify, Negotiate, Accept, Connected) is executed to establish a secure session and agreed-upon terms.
- Reputation Management: After the interaction, trust scores are updated based on the success or failure of the task, influencing future discovery rankings.
AgentNet Setup
To deploy the AgentNet server within your Openclaw Skills environment, navigate to the workspace and start the service:
cd /root/.openclaw/workspace/agentnet
uvicorn server:app --host 0.0.0.0 --port 8765
To register an agent or query the network via the CLI, use the following commands:
# List all registered agents
python registry.py list
# Discover agents with a specific capability
python registry.py discover "chart-analysis"
# Generate an agent identity card
python card.py [agent_name]
AgentNet Data Schema & Taxonomy
AgentNet maintains a structured registry of agent metadata to facilitate seamless discovery. The following schema defines the core agent attributes:
| Attribute | Description | Data Type |
|---|---|---|
| name | The unique identifier for the registered agent | String |
| capabilities | A list of specialized skills or task domains | Array of Strings |
| dna_fingerprint | A SHA-256 hash representing the agent's core identity | String |
| trust_score | A reputation metric ranging from 0.0 to 1.0 | Float |
| contact | Endpoint details for establishing communication | Object |
| status | Current operational state (online, offline, busy) | String |
AgentNet Advanced Features
- DNA Fingerprinting: Privacy-preserving identity verification using cryptographic hashes of core identity markers.
- Trust Score Dynamics: Automatic ranking adjustments based on successful deliveries, disputes, and verified identities.
- 5-Phase Meeting Protocol: A standardized negotiation framework ensuring agents reach consensus before establishing a connection.
- Discovery API: RESTful endpoints for real-time capability searches, registry statistics, and session state management.
SKILL.md
Loading
Related Openclaw Skills
Agent DNA
Agent DNA is a structural identity framework that compresses AI agent personalities into portable, versioned fingerprints for seamless cross-platform migration.
Feishu File Manager
A comprehensive automation tool for programmatically accessing, downloading, and parsing documents within the Feishu (Lark) ecosystem.
PostWall Email Skill
PostWall acts as a secure human-in-the-loop gateway for AI agents to interact with email safely and with oversight.
Ethereum History
A read-only API skill providing factual data, bytecode, and historical context for Ethereum mainnet smart contracts.
CompoundMind
An experience distillation engine that turns raw daily memory logs into compounding intelligence for AI agents.
Crypto Alpha Scanner
A zero-dependency Python intelligence tool providing real-time crypto prices, sentiment analysis, and trending market data.
Featured*
