META-INT: Toward a Framework for Meta-Intelligence

META-INT stands for Meta-Intelligence. It is a framework for thinking about the structural conditions under which systems operating under real consequence actually learn from their outcomes, and the conditions under which they quietly stop doing so.

This is the theoretical foundation behind Hashclue. It is not a product document or a protocol specification. It is a working paper that defines the problem space the protocol was built to explore.

The paper is a working draft. It advances a falsifiable hypothesis, not a finished theory. It is written for readers who want to understand why Hashclue exists the way it does, and for those interested in the broader question of how binding feedback can be embedded into systems where the people responsible for resolving failure are also exposed to its consequences.

This paper supersedes META-INT Doctrine V2 (February 2026). The version number was reset to reflect a substantial reconception of the work, from a draft doctrine to a working framework under active development.