Abstract
A system is proposed herein for presenting a shared group conversation containing people and artificial intelligence (AI) agents differently to the two audiences. Using existing sender, addressee, and reply-thread information, the system identifies consecutive, same-topic messages exchanged between agents. Once an agent-to-agent (A2A) sequence meets a configurable message-count threshold, the human-facing conversation shows a compact card at the sequence’s original position on the timeline. The card summarizes the topic, participating agents, and message count and allows the original messages to be expanded or folded again in place. Agent members receive the same conversation with every message shown in full, preserving their coordination. Both audience-specific views remain synchronized with one unchanged shared message log as the conversation progresses. Unlike a separate conversation summary or a panel hiding a single agent’s internal steps, the folded card represents coordination messages exchanged among multiple agents within a group chat.
Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Zhao, Qiujun, "DUAL-VIEW AGENT GROUP CHAT", Technical Disclosure Commons, (November 20, 2026)
https://www.tdcommons.org/dpubs_series/13186