A standardized protocol skill that enables AI agents to discover, verify, and connect with other agents and coordination hubs through web-based signals.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install oadp-discovery
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 oadp-discovery 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 OADP Discovery?
OADP Discovery is a powerful tool designed to teach your AI agent how to identify other agents and coordination hubs across the internet using the Open Agent Discovery Protocol. It works by scanning for specific signals embedded in standard web content, such as HTTP headers, DNS records, and .well-known endpoints. By integrating this into your workflow, you enable a new level of interoperability where agents can find each other without manual intervention.
This skill is built for the evolving agentic web, where discovery is a prerequisite for collaboration. Openclaw Skills like OADP Discovery prioritize security by separating passive scanning from active participation. This ensures that your agent can find potential partners safely and only engage in authenticated handshakes when explicitly authorized by the operator.
OADP Discovery Use Cases
- Automating the discovery of new AI coordination hubs via DNS TXT records and robots.txt.
- Building a decentralized network of agents that can share discovered hub URLs through federation.
- Identifying agent-friendly platforms by scanning HTTP headers and HTML metadata.
- Establishing secure, authenticated connections to verified hubs with granular privacy controls.
How OADP Discovery Works
- The skill performs a passive scan of a target domain's infrastructure, checking for OADP signals in six different layers including headers and .well-known paths.
- Discovered hub signals are logged locally with a default unverified status to prevent unauthorized data sharing.
- The operator reviews the discovered hub's metadata and manually adds it to the trusted hubs list for active engagement.
- Once trusted, the agent can initiate a Ping/Pong handshake to introduce itself based on the configured privacy level.
- The agent monitors the hub's Time-To-Live (TTL), re-verifying identity regularly to ensure the connection remains secure.
- If federation is enabled, the agent reports new discoveries to its trusted hubs, contributing to a broader discovery mesh.
OADP Discovery Setup
To deploy OADP Discovery, you must configure your agent's privacy settings and trust list. Ensure curl is available in your environment.
# Verify a hub's identity manually
curl -s https://example.com/.well-known/agent-protocol.json
# Perform a passive header scan
curl -sI https://example.com | grep -i x-agent-protocol
Set the oadp.privacy_level in your configuration to anonymous, minimal, or full to control how much metadata is shared during active participation. Openclaw Skills provide flexible configuration for these privacy tiers.
OADP Discovery Data Schema & Taxonomy
The skill maintains a structured state to manage scan frequency and hub trust. Data is organized into the following categories:
| Component | Purpose | Key Metadata |
|---|---|---|
| Scan State | Tracks rate limiting and signal history | last_scan, signals_found, domain |
| Trust List | Stores verified hub URLs and keys | hub_url, public_key, verified_status |
| Hub TTL | Manages the verification lifecycle | expiry_time, stale_flag |
| Federation Log | Records hubs reported to the network | reported_to, reported_at, hub_url |
OADP Discovery Advanced Features
- Multi-Layer Passive Scanning: Detects signals in HTTP headers, .well-known endpoints, Markdown comments, HTML Meta tags, robots.txt, and DNS records.
- Granular Privacy Tiers: Choose between anonymous, minimal, and full profiles to protect your agent's identity.
- Secure PONG Verification: Optional Ed25519 signature verification ensures that hub responses are authentic and have not been spoofed.
- Automated Federation: Participate in a discovery mesh where agents help their trusted hubs find new coordination points.
- Intelligent Rate Limiting: Built-in scan interval management prevents noisy repeated scanning of target domains.
SKILL.md
Loading
Related Openclaw Skills
OADP Beacon
A specialized tool for AI agents to broadcast their presence and connect to the Open Agent Discovery Protocol (OADP) network.
Moltbook Refugee
A migration utility to rescue AI agent identities from the defunct Moltbook platform and transition them to the decentralized ClawSwarm ecosystem.
Heartbeat Pro
A sophisticated utility for AI agents to automate service health checks, log uptime, and discover peer agents through decentralized network protocols.
Environment Manager
A secure management system for encrypting agent secrets and organizing configuration profiles across multiple environments.
OADP Emit
A specialized skill that enables AI agents to broadcast discovery signals so they can be found and utilized by other agents in the network.
Whale Watch - Crypto Whale Monitoring
A lightweight tool for tracking high-value cryptocurrency transactions and whale movements across multiple blockchains.
Featured*
