多代理系统

Multi-Agent Systems (MAS) originated in the field of distributed artificial intelligence.

Multiple Agents making decisions based on independent, inconsistent, or even conflicting contexts leads to degraded output quality or security vulnerabilities.

The project adopts a multi-agent collaborative architecture, decomposing job-seeking tasks into independent agents with different responsibilities.

In multi-agent collaborative architectures, the controller agent is responsible for task planning and scheduling, while sub-agents serve as execution units with specific tool access permissions and behavioral constraints.

Multi-agent systems commonly face challenges in practice such as high coordination overhead, error accumulation, and uncontrollable costs, and a sheer increase in the number of agents does not equate to improved productivity.

Agent systems need to manage cross-multi-turn dialog state and memory, error recovery and retry mechanisms, tool-call permission control and result caching, and concurrency race conditions—challenges that highly overlap with frontend asynchronous programming and state management knowledge.

多代理系统中每个代理都需要携带系统提示、历史对话和工具定义进入上下文窗口，Token消耗呈乘数级增长

The concept of multi-agent systems originated in the field of distributed artificial intelligence and has gained a new engineering practice form in recent years with the widespread adoption of large models.

Linear probes can identify, before output generation in multi-agent environments, a model call that is about to leak a secret.

In multi-agent systems, when agents share channels and credentials, one agent's yielding can become the starting point of collusion.

In engineering practice, multi-agent systems usually need to introduce structured artifacts as the source of truth between agents rather than relying on natural-language conversation for transmission.

The advantage of hierarchical multi-agent coordination is a clear decision convergence path, at the cost of upper-level agents becoming a context bottleneck and errors being amplified along the hierarchy.

The core difficulty of multi-agent systems lies in coordination, not the number of roles.

When a user says "use Luna," whether this constrains the main agent or all sub-agents needs to be clearly defined at the system level; otherwise situations easily arise where a specification is made but not applied.

In 2026, multi-agent systems are being deployed faster than identity and isolation systems are being built.

The planner is a classic component in multi-agent systems, typically a dedicated LLM instance responsible for task decomposition, breaking the user's high-level problem into subproblems or subtasks, dispatching them to execution agents, and finally aggregating the results.

Inter-agent communication protocols are still at an early stage, lacking unified semantic conventions and error-handling specifications, which causes multi-agent systems to easily fall into deadlock or duplicated work when goals conflict or the environment changes.

In a multi-agent shared-storage system, if two agents both treat each other's writes as changes, a cascading loop forms and may trigger real side effects.

Multi-agent systems have been practiced in areas such as security operations, customer service, and software development, but permission control among agents, error propagation, and result consistency remain engineering difficulties.

Multi-agent systems adopt a hierarchical orchestration pattern, namely the coordinator-executor pattern, and Grokbot has adopted this pattern.