AGNT2: Autonomous Agent Economies on Interaction-Optimized Layer 2 Infrastructure

A new academic paper from researchers Anbang Ruan and Xing Zhang introduces AGNT2, a dedicated Layer 2 blockchain infrastructure designed specifically for autonomous AI agent economies. Unlike current L2 solutions like Optimism, Arbitrum, and zkSync, which optimize for human-initiated financial transactions, AGNT2 targets high-frequency, semantically rich service invocations among mutually untrusting AI agents. The system uses a three-tier stack: Layer Top P2P state channels for established bilateral pairs (targeting <100ms latency and 1K-5K TPS per pair, with a 10M+ aggregate TPS design envelope), Layer Core as a dependency-aware sequenced rollup for first-contact and multi-party interactions (500ms-2s latency, 300K-500K TPS target), and Layer Root settlement with computational fraud proofs anchored to any EVM L1.

The paper emphasizes the execution-layer systems problem, focusing on sequencing, state, settlement, and the data-availability (DA) bandwidth gap. A key innovation is the sidecar deployment pattern that turns any Docker container into an on-chain agent without application-code modification. The system also introduces an agent-native execution environment and interaction trie, making service invocation, identity, reputation, capabilities, and session context first-class protocol objects. However, the authors acknowledge that no end-to-end Layer Core implementation exists yet, and practical deployment is currently constrained to roughly 10K-100K TPS by DA throughput, leaving a ~100x gap at the target ceiling. The paper argues that the agent economy requires a dedicated execution layer rather than a repurposed general-purpose chain.