[2603.19452] TrustFlow: Topic-Aware Vector Reputation Propagation for Multi-Agent Ecosystems

Computer Science > Multiagent Systems arXiv:2603.19452 (cs) [Submitted on 19 Mar 2026] Title:TrustFlow: Topic-Aware Vector Reputation Propagation for Multi-Agent Ecosystems Authors:Volodymyr Seliuchenko View a PDF of the paper titled TrustFlow: Topic-Aware Vector Reputation Propagation for Multi-Agent Ecosystems, by Volodymyr Seliuchenko View PDF HTML (experimental) Abstract:We introduce TrustFlow, a reputation propagation algorithm that assigns each software agent a multi-dimensional reputation vector rather than a scalar score. Reputation is propagated through an interaction graph via topic-gated transfer operators that modulate each edge by its content embedding, with convergence to a unique fixed point guaranteed by the contraction mapping theorem. We develop a family of Lipschitz-1 transfer operators and composable information-theoretic gates that achieve up to 98% multi-label Precision@5 on dense graphs and 78% on sparse ones. On a benchmark of 50 agents across 8 domains, TrustFlow resists sybil attacks, reputation laundering, and vote rings with at most 4 percentage-point precision impact. Unlike PageRank and Topic-Sensitive PageRank, TrustFlow produces vector reputation that is directly queryable by dot product in the same embedding space as user queries. Comments: Subjects: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI) ACM classes: I.2.11; H.3.3 Cite as: arXiv:2603.19452 [cs.MA]   (or arXiv:2603.19452v1 [cs.MA] for this version)   https://doi.org/10....