Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.15034 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909054781095936 |
|---|---|
| author | Zhang, Wentao Zhao, Zhe Wen, Haibin Wu, Yingcheng Guo, Cankun Yin, Ming An, Bo Wang, Mengdi |
| author_facet | Zhang, Wentao Zhao, Zhe Wen, Haibin Wu, Yingcheng Guo, Cankun Yin, Ming An, Bo Wang, Mengdi |
| contents | Recent advances in LLM based agent systems have shown promise in tackling complex, long horizon tasks. However, existing agent protocols (e.g., A2A and MCP) under specify cross entity lifecycle and context management, version tracking, and evolution safe update interfaces, which encourages monolithic compositions and brittle glue code. We introduce Autogenesis Protocol (AGP), a self evolution protocol that decouples what evolves from how evolution occurs. Its Resource Substrate Protocol Layer (RSPL) models prompts, agents, tools, environments, and memory as protocol registered resources with explicit state, lifecycle, and versioned interfaces. Its Self Evolution Protocol Layer (SEPL) specifies a closed loop operator interface for proposing, assessing, and committing improvements with auditable lineage and rollback. Building on AGP, we present Autogenesis System (AGS), a self-evolving multi-agent system that dynamically instantiates, retrieves, and refines protocol-registered resources during execution. We evaluate AGS on multiple challenging benchmarks that require long horizon planning and tool use across heterogeneous resources. The results demonstrate consistent improvements over strong baselines, supporting the effectiveness of agent resource management and closed loop self evolution. The code is available at https://github.com/DVampire/Autogenesis. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_15034 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Autogenesis: A Self-Evolving Agent Protocol Zhang, Wentao Zhao, Zhe Wen, Haibin Wu, Yingcheng Guo, Cankun Yin, Ming An, Bo Wang, Mengdi Artificial Intelligence Recent advances in LLM based agent systems have shown promise in tackling complex, long horizon tasks. However, existing agent protocols (e.g., A2A and MCP) under specify cross entity lifecycle and context management, version tracking, and evolution safe update interfaces, which encourages monolithic compositions and brittle glue code. We introduce Autogenesis Protocol (AGP), a self evolution protocol that decouples what evolves from how evolution occurs. Its Resource Substrate Protocol Layer (RSPL) models prompts, agents, tools, environments, and memory as protocol registered resources with explicit state, lifecycle, and versioned interfaces. Its Self Evolution Protocol Layer (SEPL) specifies a closed loop operator interface for proposing, assessing, and committing improvements with auditable lineage and rollback. Building on AGP, we present Autogenesis System (AGS), a self-evolving multi-agent system that dynamically instantiates, retrieves, and refines protocol-registered resources during execution. We evaluate AGS on multiple challenging benchmarks that require long horizon planning and tool use across heterogeneous resources. The results demonstrate consistent improvements over strong baselines, supporting the effectiveness of agent resource management and closed loop self evolution. The code is available at https://github.com/DVampire/Autogenesis. |
| title | Autogenesis: A Self-Evolving Agent Protocol |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2604.15034 |