| International Journal of Computer Applications |
| Foundation of Computer Science (FCS), NY, USA |
| Volume 187 - Number 84 |
| Year of Publication: 2026 |
| Authors: Prudhvi Ratna Badri Satya, Ajay Guyyala, Krishna Teja Areti, Vijay Putta |
10.5120/ijca2026926474
|
Prudhvi Ratna Badri Satya, Ajay Guyyala, Krishna Teja Areti, Vijay Putta . Cross-Server Interoperability in Multi-MCP Automated AI Agent Networks. International Journal of Computer Applications. 187, 84 ( Feb 2026), 1-14. DOI=10.5120/ijca2026926474
This paper introduced a combined framework for cross-server interoperability in multi-MCP automated AI agent networks. The design combined communication abstraction, orchestration optimization, and security validation. The framework was tested on BoT-IoT, ToN-IoT, and PettingZoo datasets, which represented adversarial traffic detection, telemetry-heavy IoT environments, and dynamic multi-agent orchestration. Results showed improvements in coverage, efficiency, and robustness, with accuracy, precision, recall, and F1-score above 0.95 across multiple trials. Ablation analysis confirmed the role of each component, scalability tests showed stable performance as servers increased, and stress evaluations demonstrated graceful degradation under heavier attack loads. Error analysis and statistical validation supported the reliability of the outcomes, while resource usage comparisons indicated reduced runtime and memory consumption against baselines. Cross-domain generalization confirmed adaptability across unseen datasets. These findings demonstrated that interoperability in heterogeneous MCP networks can be achieved without sacrificing efficiency, scalability, or reliability, providing a foundation for secure and practical multi-domain agent collaboration.