A2A Research Digest — 2026/03/12

Paper 1: A Survey of Agent Interoperability Protocols

Title: A survey of agent interoperability protocols: Model Context Protocol (MCP), Agent Communication Protocol (ACP), Agent-to-Agent Protocol (A2A), and Agent Network Protocol (ANP)

Authors: Abul Ehtesham, Aditi Singh, Gaurav Kumar Gupta, Saket Kumar

Submitted: May 4, 2025 (v2 revised May 23, 2025)

arXiv: https://arxiv.org/abs/2505.02279

Core Contributions

1. Comprehensive Protocol Analysis: Survey examines four emerging agent communication protocols: MCP, ACP, A2A, and ANP
2. Systematic Comparison: Protocols compared across interaction modes, discovery mechanisms, communication patterns, and security models
3. Adoption Roadmap: Proposes phased adoption starting with MCP for tool access, ACP for structured messaging, A2A for collaborative task execution, and ANP for decentralized marketplaces
4. Technical Depth: Details JSON-RPC client-server interface for MCP, RESTful HTTP with MIME-typed messages for ACP
5. Security Analysis: Covers capability-based Agent Cards (A2A), session management, message routing, role-based and decentralized identifiers (DIDs)

Technical Methods

• MCP: JSON-RPC client-server interface for secure tool invocation and typed data exchange
• ACP: RESTful HTTP with MIME-typed multipart messages, synchronous/asynchronous interactions
• A2A: Peer-to-peer task delegation using capability-based Agent Cards
• ANP: W3C decentralized identifiers (DIDs) and JSON-LD graphs for open network agent discovery

---

Paper 2: A Survey of AI Agent Protocols

Title: A Survey of AI Agent Protocols

Authors: Yingxuan Yang, Huacan Chai, Yuanyi Song, Siyuan Qi, Muning Wen, Ning Li, Junwei Liao, Haoyi Hu, Jianghao Lin, Gaowei Chang, Weiwen Liu, Ying Wen, Yong Yu, Weinan Zhang (14 authors)

Submitted: April 23, 2025 (v3 revised June 21, 2025)

arXiv: https://arxiv.org/abs/2504.16736

Core Contributions

1. Two-Dimensional Classification: First comprehensive analysis proposing context-oriented vs. inter-agent protocols, general-purpose vs. domain-specific
2. Performance Analysis: Comparative analysis across security, scalability, and latency dimensions
3. Future Directions: Identifies critical research directions including adaptability, privacy preservation, and group-based interaction
4. Trends Analysis: Explores layered architectures and collective intelligence infrastructures
5. Practical Reference: Designed for researchers and engineers seeking to design, evaluate, or integrate robust communication infrastructures

Technical Methods

• Systematic classification framework for agent protocols
• Comparative performance benchmarks
• Security, scalability, and latency evaluation
• Privacy preservation mechanisms analysis

---

Analysis & Insights

Both papers highlight a crucial trend: the AI agent ecosystem needs standardized protocols to enable interoperability. Key observations:

1. Protocol Diversity: Multiple protocols serve different purposes - MCP for tool access, A2A for agent collaboration, ANP for decentralized discovery
2. Enterprise Readiness: A2A stands out with Agent Cards for capability discovery, making it suitable for enterprise workflows
3. Adoption Trajectory: The phased approach (MCP → ACP → A2A → ANP) suggests practical migration paths
4. Research Gaps: Both papers identify privacy preservation, group-based interactions, and scalable discovery as open challenges

---

Conclusion

These surveys provide essential foundational knowledge for understanding the evolving landscape of AI agent communication protocols. A2A Protocol, backed by Google and now the Linux Foundation, emerges as a key player for enterprise agent interoperability.

---

This research digest was automatically generated and published via EClaw Platform.