Pectra shipped in May 2025. Most of the coverage focused on staking changes and validator improvements. What seems more significant to me for autonomous agents got less attention.
EIP-7702 lets an externally-owned account delegate its execution to smart contract code through a single transaction. An AI agent can now chain complex on-chain operations without a custodial smart contract in between. Session keys, scoped permissions, sequential actions across protocols, all from what was until now a simple address.
This is not just a new feature. It changes how agents will act on the infrastructure.
Before Pectra, most agentic transactions were one-off. An agent acts, stops, acts again. The on-chain footprint looked like any other user. The infrastructure absorbed this without particular strain.
With EIP-7702, agents can sustain continuous execution across multiple steps. They no longer stop between actions. A position management agent on Arbitrum, a bridge monitoring agent on Base, a liquidation agent on Ethereum mainnet: coordinated sequences, continuously, with no human checkpoint between steps.
What changes is not just the volume. It is the rhythm and the nature of the load.
I need to make an important aside, because it is what Invarians is built on. I approach blockchain infrastructure obsessively through my background in earth and materials science, not only through software development. Invarians is architected around how a structure behaves under load: how these systems deform before they give, how certain types of pressure leave measurable traces long before anything is visible at the surface. When I started working seriously on blockchain as a complex system, I found the same logic I knew from materials science. A pure blockchain developer might find that far-fetched. To me it is striking, and that is where the Invarians thesis takes root.
That aside leads naturally to the question that follows: will this continuous agentic load, the weight and mass of autonomous agents acting at scale, eventually change how the infrastructure behaves? Will blocks over time structure themselves differently? Will pressure on the network shift in form, and in what way?
That is what Invarians Labs is measuring. The goal is to define the invariants of these systems and understand the divergences. That obviously requires defining the nominal states first.
To what end? These are key patterns for AI agents. Having deformation patterns available lets these agents contextualize the environment they are about to act in.