SBKIM: A Protocol for Semantic Bidirectional AI Matching in Human and Agent Networks

Abstract

We propose SBKIM (Semantic Bidirectional AI Matching), an open protocol fordetermining semantic compatibility between two parties in natural language,without structured queries, databases or centralised platforms. Unlikeunidirectional retrieval systems, SBKIM requires that both parties describetheir capabilities and their needs at the same time. A language model ratesthe pair on three orthogonal dimensions and returns a structuredcompatibility report. The protocol operates on two layers: layer 1 forhuman-to-human matching (buyer/supplier, employee/project) and layer 2 foragent-to-agent matching in multi-agent systems. This is a field report, not a novelty claim: the building blocks are known,the operation is the contribution. A working reference implementation ispublicly available, and the traffic of the running network can be observedthrough a read-only network map. The German version of this paper is included as a second file. Reference implementation: https://github.com/lausiklauskn-png/Sage-Protokol