Table of Contents
Fetching ...

Infrastructure for Valuable, Tradable, and Verifiable Agent Memory

Mengyuan Li, Lei Gao, Haoxuan Xu, Jiate Li, Potung Yu, Lingke Cheng, Yue Zhao, Murali Annavaram

Abstract

Every API token you spend is your accumulated wealth; once you can prove its value and the effort behind it, you can resell it. As autonomous agents repeatedly call models and tools, they accumulate memories that are your intellectual property. But today these memories remain private and non-transferable, as there is no way to validate their value. We argue that agent memory can serve as an economic commodity in the agent economy, if buyers can verify that it is authentic, effort-backed, and produced in a compatible execution context. To realize this idea, we propose clawgang, which binds memory to verifiable computational provenance, and meowtrade, a market layer for listing, transferring, and governing certified memory artifacts. Together, they transform one-shot API token spending into reusable and tradable assets, enabling timely memory transfer, reducing repeated exploration, and opening a memory trade market.

Infrastructure for Valuable, Tradable, and Verifiable Agent Memory

Abstract

Every API token you spend is your accumulated wealth; once you can prove its value and the effort behind it, you can resell it. As autonomous agents repeatedly call models and tools, they accumulate memories that are your intellectual property. But today these memories remain private and non-transferable, as there is no way to validate their value. We argue that agent memory can serve as an economic commodity in the agent economy, if buyers can verify that it is authentic, effort-backed, and produced in a compatible execution context. To realize this idea, we propose clawgang, which binds memory to verifiable computational provenance, and meowtrade, a market layer for listing, transferring, and governing certified memory artifacts. Together, they transform one-shot API token spending into reusable and tradable assets, enabling timely memory transfer, reducing repeated exploration, and opening a memory trade market.

Paper Structure

This paper contains 13 sections, 3 figures.

Figures (3)

  • Figure 1: A general workflow for gang creation and member registration.
  • Figure 2: System overview of certified memory accumulation across VMPL0 and VMPL3. VMPL3 hosts the agent runtime and working memory, while VMPL0 is the trusted domain responsible for provider authentication, prompt/response measurement and anchored-root accumulation.
  • Figure 3: Selective disclosure over a measured interaction log. TEE/VMPL0 measures each prompt-response pair into per-field digests and then into a per-interaction digest, while continuously accumulating an anchored root over the interaction sequence. During trade, selected fields may be disclosed in plaintext and hidden fields only as digests, enabling the buyer to verify consistency with the anchored root without learning undisclosed content.