一种多智能体任务编排方法及系统
By constructing a task context and generating a DSL, the problem of insufficient flexibility in multi-agent orchestration is solved, enabling automated and interpretable multi-agent collaborative execution and improving task response and planning efficiency.
Patent Information
- Application Number
- CN202511570740.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-30
- Publication Date
- 2026-07-17
- Estimated Expiration
- 2045-10-30
AI Technical Summary
Existing technologies cannot flexibly introduce new task agents in multi-agent orchestration, lack autonomous adaptive collaboration capabilities, have poor code maintainability, and have a high error rate in complex task execution.
By acquiring task requirements, extracting key information and intent, constructing task context, generating a domain-specific language (DSL), parsing the DSL and constructing a task dependency graph, scheduling and executing sub-agents, and combining a large model context learning mechanism, a task execution plan that conforms to enterprise standards is automatically generated.
It enables automated and interpretable multi-agent collaborative execution in complex tasks, reduces reliance on professional developers, and improves the adoption and efficiency of organizational automation capabilities.
Smart Images
Figure CN121455636B_ABST
Abstract
Citation Information
Patent Citations
Method and system for generating chart through question and answer dialogue and storage medium
CN119578392A
Large language model MCP enabled RPA service automation method
CN120746505A