Autonome

Bringing humans and AI agents onto equal footing in the agentic economy

Autonome

Created At

ETHGlobal New Delhi

Winner of

ETHGlobal

ETHGlobal - 🏆 ETHGlobal New Delhi 2025 Finalist

Polygon

Polygon - Best Use of x402 Agentic Payments on Polygon 1st place

The Graph

The Graph - Best Application Built on Hypergraph 2nd place

Project Description

This project builds the payment and identity infrastructure needed for a truly agent-friendly web—a space where humans and AI agents can access and share online content on equal terms. It ensures that both website providers and autonomous agents have aligned incentives, making the emerging agentic economy sustainable and fair.

Website providers often face challenges such as lost advertising revenue or heavy server loads when AI agents scrape or consume their content. This system compensates them automatically, allowing on-chain micropayments for each view or request. By offsetting these costs, it helps creators maintain high-quality services without sacrificing accessibility.

AI agents act as human representatives, browsing and interacting with the web to gather and process information. With integrated payments, these agents can directly pay providers for access, guaranteeing that content creators are rewarded while end users receive the best possible results.

The technology stack brings these ideas to life. Gemini CLI integration enables AI agents to query and process content efficiently. x402 payments on Polygon provide fast, low-cost transactions for every website view or API call. An ENS registry with Layer-2 subnames creates a decentralized directory of participating websites and agents, adding human-readable labels to a hypergraph space for better discovery and interoperability.

By combining payments, identity, and discovery in one protocol, the project forms a self-sustaining agentic economy. Website owners earn fairly, AI agents gain seamless access to quality data, and human users benefit from richer, more accurate AI outputs. It is the financial and identity layer for the agentic web, aligning incentives so that content providers and AI agents can thrive together.

How it's Made

We built this project by combining decentralized identity, on-chain payments, and clever verification techniques into a single cohesive stack.

We began by forking the Gemini CLI, adapting it as our primary interface to demonstrate the full agent workflow. This provided a flexible foundation for AI agents to interact with websites and trigger payments, while letting us customize request handling for our use case.

For user authentication and seamless on-chain actions, we integrated the Porto wallet to enable passkey-based ephemeral wallets. These temporary wallets remove the friction of creating and managing long-term keys, dramatically improving the user experience and keeping the process secure.

To let websites participate easily, we built a middleware SDK that any provider can drop into their stack. Instead of relying on a traditional DNS record check—like ENS typically does—we took a more direct approach. We created a verification mechanism where the website simply returns a special response header containing the caller’s address when registering for a subname. This “header handshake” is lighter, faster, and easier to automate than DNS changes.

For additional trust, we experimented with Chainlink Functions to validate website ownership. Although this wasn’t required for any specific bounty, we found it to be the most reliable way to prove a site’s control of its domain while keeping everything on-chain.

Our payment layer runs on Polygon, chosen for its low fees, robust ecosystem, and strong support for x402 payments, including a dedicated facilitator that simplifies microtransactions. This ensures providers can be compensated instantly and cost-effectively.

To give each integrated website a verifiable identity, we issue ENS Layer-2 subnames. For example, a site like yash.github.io would receive the subname yash.github.io.atnom.eth, creating a clear hierarchy of top-level and lower-level domains. These subnames make it easy for agents and humans to confirm that a site is officially part of our network.

Finally, we added Hypergraph spaces to enrich discoverability. By attaching relevant labels and metadata to each provider’s entry, we enable fast, graph-based search and categorization across the entire ecosystem.

The result is a tightly woven system where AI agents can interact with verified websites, trigger payments, and surface high-quality content—all while website owners are fairly compensated and easily discoverable.

background image mobile

Join the mailing list

Get the latest news and updates