A task processing method and device, a storage medium and a computer program product

Abstract

The application provides a task processing method and device, a storage medium and a computer program product. The method comprises the following steps: receiving a big data development task, and matching a first intelligent agent corresponding to the big data development task; acquiring a first tool corresponding to the first intelligent agent; accessing a large model through a large model adaptation service, and selecting a target tool for the big data development task from the first tool by using the large model; and calling a big data platform to execute the big data development task by using the target tool.

Description

Technical Field

[0001] This application relates to electronic application technology, and more particularly to a task processing method and apparatus, storage medium, and computer program product. Background Technology

[0002] Data development refers to the entire process of collecting, storing, processing, analyzing, and applying massive, diverse, and rapidly generated data using big data technologies. Its core objective is to build efficient and reliable big data systems to support business decision-making, data mining, and intelligent applications. Currently, data development can be based on large models, primarily using natural language to generate Structured Query Language (SQL), and then extending these models to create dedicated intelligent agents to achieve the aforementioned functions. However, this data development process is functionally limited and cannot meet diverse data development needs; furthermore, the tight coupling between the large model and the intelligent agent leads to poor compatibility between the agent and the large model. Summary of the Invention

[0003] This application provides a task processing method and apparatus, a storage medium, and a computer program product.

[0004] The technical solution of this application is implemented as follows: Firstly, this application proposes a task processing method, the method comprising: Receive big data development tasks and match the first intelligent agent corresponding to the big data development tasks; Obtain the first tool corresponding to the first intelligent agent; and access the large model through the large model adaptation service, and use the large model to select the target tool for the big data development task from the first tool; The target tool is used to invoke the big data platform to execute the big data development task.

[0005] Secondly, this application proposes a task processing apparatus, the apparatus comprising: The receiving unit is used to receive big data development tasks; The matching unit is used to match the first intelligent agent corresponding to the big data development task; The acquisition unit is used to acquire the first tool corresponding to the first intelligent agent; The selection unit is used to access the large model through the large model adaptation service and use the large model to select the target tool for the big data development task from the first tool. The calling unit is used to invoke the big data platform using the target tool to execute the big data development task.

[0006] Thirdly, this application proposes a task processing apparatus, the apparatus comprising: a processor, a memory, and a communication bus; the processor executes a running program stored in the memory to implement the above-mentioned task processing method.

[0007] Fourthly, this application provides a storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described task processing method.

[0008] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the above-described task processing method.

[0009] The embodiments of this application have the following beneficial effects: the modular design can support the expansion of different types of intelligent agents to be compatible with big data development tasks with different needs, and can meet diverse data development needs; by accessing different large models through the large model adaptation service, the large model and the intelligent agent are decoupled, and the upgrade, update, failure and other situations of the large model will not affect the performance of the intelligent agent, thereby improving the compatibility of the intelligent agent with the large model. Attached Figure Description

[0010] Figure 1 A flowchart illustrating a task processing method provided in an embodiment of this application; Figure 2 A connection diagram illustrating an exemplary artificial intelligence (AI) agent routing service provided in this application embodiment; Figure 3 A schematic diagram illustrating the connection of an exemplary Large Language Model (LLM) adapter provided in this application embodiment; Figure 4 A schematic diagram of an exemplary system architecture for big data task development based on intelligent agents, provided for embodiments of this application; Figure 5 This application provides an exemplary flowchart for developing big data tasks based on intelligent agents. Figure 6 A schematic diagram of the structure of a task processing device provided in this application embodiment. Figure 1 ; Figure 7 A schematic diagram of the structure of a task processing device provided in this application embodiment. Figure 2 .

[0011] It should be noted that the terms "first" and "second" mentioned above are only used to distinguish between different options and do not represent the degree of superiority or inferiority of the options or their priority in the implementation process. Detailed Implementation

[0012] In order to gain a more detailed understanding of the features and technical content of the embodiments of this application, the implementation of the embodiments of this application will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for reference and illustration only and are not intended to limit the embodiments of this application.

[0013] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0014] In the following description, references to "some embodiments" refer to a subset of all possible embodiments. It is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. It should also be noted that the terms "first, second, third" used in the embodiments of this application are merely for distinguishing similar objects and do not represent a specific ordering of objects. It is understood that "first, second, third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0015] This application provides a task processing method, such as... Figure 1 As shown, the method may include: S101. Receive big data development tasks and match the first intelligent agent corresponding to the big data development tasks.

[0016] In this embodiment of the application, the big data development tasks sent by various clients are received through various interfaces provided by the intelligent agent routing service, and the first intelligent agent corresponding to the big data development task is selected.

[0017] It should be noted that the intelligent agent routing service can be an AI intelligent agent routing service, which is a gateway layer that provides various interfaces to connect to various clients.

[0018] Optionally, the various interfaces may include: Hyper Text Transfer Protocol (HTTP), Secure File Transfer Protocol (SFTP), Transmission Control Protocol (TCP), Message Queuing Telemetry Transport (MQTT), etc. The specific type can be selected according to the actual situation, and this application embodiment does not make specific limitations.

[0019] Optionally, various types of clients may include: World Wide Web (Web) applications, mobile apps, Internet of Things (IoT) devices, smart home systems, etc. The specific types can be selected according to the actual situation, and this application embodiment does not impose specific limitations.

[0020] It should be noted that the AI ​​intelligent agent routing service provides interface adaptation capabilities to meet the needs of different clients.

[0021] In this embodiment, the AI ​​agent routing service also provides capabilities such as load balancing and traffic control, enabling the agent to be freed from input layer tasks and focus more on improving its core capabilities. The agent can then dedicate more resources and energy to algorithm optimization, model training, and knowledge reasoning, continuously improving its intelligence level and processing power.

[0022] It should be noted that the intelligent agent is an AI intelligent agent.

[0023] It should be noted that the AI ​​intelligent agent routing service can match the first intelligent agent corresponding to the big data development task based on the component type of big data.

[0024] Optionally, the first intelligent agent can be an ETL agent, a Flink agent, a Spark agent, etc. The specific agent can be selected according to the actual situation, and this application embodiment does not impose specific limitations.

[0025] For example, see Figure 2 The AI ​​agent routing service connects to the client on one end and to (Extract, Transform, Load, ETL) agents, Flink agents, and Spark agents on the other end. Upon receiving a big data development task, the AI ​​agent routing service matches the task with the corresponding first agent and transmits the task to that agent.

[0026] S102. Obtain the first tool corresponding to the first intelligent agent; and access the large model through the large model adaptation service, and use the large model to select the target tool for the big data development task from the first tool.

[0027] In this embodiment of the application, the capabilities of the big data platform are integrated into multiple types of tools, forming a Model Context Protocol (MCP) service corresponding to each type of tool; then, the first tool is obtained from the first MCP service corresponding to the first intelligent agent.

[0028] For example, an ETL MCP service may include tools for data crawling, transformation, and writing; a Spark MCP service may include tools corresponding to the Spark computing engine; and a Flink MCP service may include tools corresponding to the Flink computing engine.

[0029] In one embodiment, the MCP service can be omitted, and the tool can be invoked directly through the function call method provided by the large model. The specific choice can be made according to the actual situation, and this application does not impose specific limitations.

[0030] It should be noted that the first intelligent agent can obtain the functions of the first MCP service through the standard MCP protocol.

[0031] Optionally, ETL agents correspond to ETL MCP services, Flink agents correspond to Flink MCP services, and Spark agents correspond to Spark MCP services.

[0032] It's important to note that the MCP service acts as a bridge between the big data platform and the AI ​​agent. On one hand, the MCP service possesses powerful functional integration and transformation capabilities. It can delve into the big data platform, sorting and analyzing its various functions, integrating them into a series of tools, and providing natural language descriptions. This makes abstract and logically complex big data functions clear, intuitive, and easy to understand, significantly lowering the barrier to entry and laying a solid foundation for widespread application in related fields. On the other hand, the MCP service is also used for information transmission. It provides the integrated and described big data functional information to the AI ​​agent. This allows the AI ​​agent to fully and deeply understand the capabilities of the big data platform, enabling it to better collaborate with the platform, achieving more efficient and intelligent data processing and analysis, and providing strong support for the efficient operation of the entire system.

[0033] Understandably, the intelligent agents and their corresponding MCP services adopt a plug-in design, avoiding intrusion into the existing platform. Like independent "functional modules," they can be easily integrated into the existing platform without modifying its core code, minimizing the impact on existing business operations and ensuring system stability and reliability. Each intelligent agent or MCP service implements only its corresponding function. On one hand, this single responsibility principle makes the functions of each system component clearer and simpler. On the other hand, this high degree of functional focus not only improves the performance and efficiency of each component but also makes system maintenance and upgrades easier. When a functional module malfunctions, only that module needs to be repaired or replaced without affecting the normal operation of other modules.

[0034] In this embodiment, a first intelligent agent is used to analyze big data development tasks in conjunction with historical development tasks to obtain task requirements; the task requirement information is analyzed using a large model, and a target tool is selected from the first tool.

[0035] It should be noted that the first intelligent agent can perform contextual analysis based on the interaction data of historical development tasks to optimize task interpretation and obtain task requirements.

[0036] In one embodiment, by modifying some MCP services and using similar descriptions, the large model can complete the corresponding parsing.

[0037] In this embodiment, the first intelligent agent interacts with the large model through the large model adaptation service, carrying the first tool and task requirements. The large model can parse the task requirement information and select the target tool for the big data development task from the first tool.

[0038] In an alternative embodiment, the interaction between the first agent and the large model can also be achieved through other means, such as coding.

[0039] In this embodiment, the large model adaptation service can be an LLM large model adaptation service.

[0040] It should be noted that the first agent can use the function call method to access the large model through the LLM large model adaptation service.

[0041] It's important to note that while the OpenAI protocol is mainstream, open-source large models exhibit significant differences in data format, interface specifications, and invocation methods. The LLM Large Model Adaptation Service, through research into the characteristics of various large models, has established a universal adaptation framework. This framework primarily includes standardizing request method parameters, transforming the "language" of different large models into a unified format that agents can understand. Regardless of the protocol or architecture of the large model, the LLM Large Model Adaptation Service can parse and convert it, allowing agents to access the capabilities of various large models without needing to concern themselves with the underlying implementation details. Agents can simply interact with the adaptation service according to the unified interface specifications to invoke the capabilities of various large models.

[0042] Understandably, the large model adaptation service, acting as a bridge between agents and large models, possesses strong decoupling capabilities. In traditional system architectures, agents are deeply bound to specific large models. Once the large model is upgraded or replaced, the agent often requires large-scale code modifications and functional adjustments. This not only increases development and maintenance costs but can also lead to system instability during the transition period. The large model adaptation service, however, abstracts and encapsulates the interaction between agents and large models by establishing a standardized set of interfaces and protocols. Agents only need to communicate with the adaptation service through these standard interfaces, without directly interacting with the underlying large model. Thus, even if the underlying large model changes, as long as the large model adaptation service can adapt according to the standard interfaces, the agent does not need to be modified, thereby achieving decoupling between the agent and the large model.

[0043] Optionally, the large model may include a language large model, a multimodal large model, a code large model, a professional domain large model, etc. The specific model can be selected according to the actual situation, and the embodiments of this application do not impose specific limitations.

[0044] For example, see Figure 3 The LLM large model adapter can connect to ETL agents, Flink agents, Spark agents, as well as large language models, multimodal models, code models, and domain-specific models.

[0045] S103. Use the target tool to call the big data platform to execute big data development tasks.

[0046] In this embodiment of the application, the target tool is used to call the first MCP service corresponding to the first intelligent agent; the first MCP service is used to call the big data platform to execute big data development tasks.

[0047] Furthermore, the system obtains the first result information of the big data platform executing big data development tasks; the first result information is machine language result information; it accesses the big model through the big model adaptation service, and uses the big model to convert the first result information into second result information; the second result information is natural language result information; and sends the second result information.

[0048] For example, if you input the big data development task "Please help me create a shell task named test_ai_shell", the returned second result information may include the shell task name, node, creator, etc. The specific details can be selected according to the actual situation. This application is just an example and does not impose specific limitations.

[0049] It should be noted that contingency plans are configured for at least one of the following abnormal scenarios: agent matching failure, large model access failure, and big data platform call timeout. These contingency plans include, but are not limited to, at least one of the following: retry, alarm, backup switching, and logging. The specific choice can be made based on the actual situation, and this application embodiment does not impose any specific limitations.

[0050] Understandably, adopting a modular design can support the expansion of different types of intelligent agents to be compatible with big data development tasks with different needs, and can meet diverse data development requirements; by accessing different large models through large model adaptation services, the large model and intelligent agent are decoupled, and the upgrades, updates, and failures of the large model will not affect the performance of the intelligent agent, thus improving the compatibility of the intelligent agent with the large model.

[0051] Based on the above embodiments, a system for developing big data tasks based on intelligent agents is proposed. See the system architecture diagram below. Figure 4 This includes AI agent routing services, LLM large model adapters, AI agents, and MCP services. AI agents include ETL agents, Spark agents, and Flink agents; MCP services include ETL MCP services, SparkMCP services, and Flink MCP services. The AI ​​agent routing service receives natural language requests from clients, selects the appropriate AI agent, and the AI ​​agent, carrying tools provided by the MCP service, interacts with the large model through the LLM model adapter. Then, the tool selected by the large model calls the corresponding MCP server, which in turn calls the big data platform to complete the task development. The large model can include language large models, multimodal large models, code large models, and domain-specific large models. The big data platform consists of a big data development platform, ETL tasks, Spark tasks, Flink tasks, and other task types.

[0052] This paper proposes a method for big data development using intelligent agents. Through a unified intelligent agent routing service, various intelligent agents are accessed. These agents, combined with the MCP service, interact with large models through large model adaptation services to fulfill user requirements. No complex operations or understanding of the entire system are required; complex data development and understanding of the system from intelligent agents can be completed simply through natural language, significantly reducing usage costs. Furthermore, it automatically utilizes existing platform capabilities to assist in task development, reducing workloads from hours to minutes, greatly improving efficiency. The modular and plug-in design supports flexible expansion with different types of intelligent agents to accommodate diverse needs, and adapts to various large models through the large model adaptation service, demonstrating excellent compatibility and forward-looking capabilities.

[0053] Based on the above embodiments, a method for task creation based on intelligent agents is proposed, see [link to relevant documentation]. Figure 5 The method includes: 1. The AI ​​intelligent agent routing service receives the creation task sent by the client: "Create a shell-type task named 'shell' in the root directory, with the script 'echo hello world'."

[0054] 2. The AI ​​agent routing service parses and creates tasks, and matches the AI ​​agent corresponding to the created task.

[0055] 3. The AI ​​agent routing service sends corresponding task processing instructions to the AI ​​agent.

[0056] 4. The AI ​​agent obtains a list of MCP tools for creating tasks from the MCP service.

[0057] 5. The AI ​​agent calls the large model through the LLM large model adaptation service and sends the creation task content and MCP tool list to the large model.

[0058] 6. The large model parses the task content and extracts keywords from it: Name: shell, Type: shell, Script: echo hello world. Then, select MCP tool 1 from the MCP tool list.

[0059] 7. The large model returns MCP tool 1 to the AI ​​agent through the LLM large model adaptation service.

[0060] 8. The AI ​​agent carries the creation task and calls MCP tool 1, which in turn calls the big data platform interface.

[0061] 9. The big data platform executes the creation task and generates complete configuration information for the creation task.

[0062] 10. The big data platform returns complete configuration information for the task creation to the AI ​​agent through Tool 1.

[0063] It should be noted that the complete configuration information for creating a task can include all content such as the task's basic identifier, its directory and project space, the core execution script, task attributes, scheduling strategy, debugging configuration, creation and update information, etc. The task type is a Shell script, it is currently offline, the scheduling rule is to execute it every 5 minutes, the timeout is 1 hour, and it will retry once on failure. The content of the execution script is to output "hello world", and it also includes related information such as request ID, status code, execution resources, and responsible person, which fully records the configuration and metadata of the created task.

[0064] 11. The AI ​​agent sends the complete configuration information of the creation task to the large model through the LLM large model adaptation service.

[0065] 12. The large model integrates the complete configuration information of the creation task into a natural language result document.

[0066] 13. The large model sends the natural language result text to the AI ​​agent through the LLM large model adaptation service.

[0067] 14. The AI ​​agent sends natural language results to the client through the AI ​​agent routing service.

[0068] This application provides a task processing device. For example... Figure 6 As shown, the task processing device 1 includes: Receiving unit 10 is used to receive big data development tasks; Matching unit 11 is used to match the first intelligent agent corresponding to the big data development task; Acquisition unit 12 is used to acquire the first tool corresponding to the first intelligent agent; Selection unit 13 is used to access a large model through the large model adaptation service and use the large model to select a target tool for the big data development task from the first tool. Calling unit 14 is used to call the big data platform using the target tool to execute the big data development task.

[0069] Optionally, the receiving unit 10 is further configured to receive big data development tasks sent by corresponding clients through multiple interfaces provided by the intelligent agent routing service; The matching unit 11 is also used to match the first intelligent agent corresponding to the big data development task.

[0070] Optionally, the calling unit 14 is further configured to use the target tool to call the first model context protocol (MCP) service corresponding to the first intelligent agent; The first MCP service is used to call the big data platform to execute the big data development task.

[0071] Optionally, the device further includes: a conversion unit and a transmission unit; The acquisition unit 12 is further configured to acquire first result information of the big data platform executing the big data development task; the first result information is machine language result information; The conversion unit is used to access the large model through the large model adaptation service and use the large model to convert the first result information into second result information; the second result information is natural language result information. The sending unit is used to send the second result information.

[0072] Optionally, the device further includes: an integration unit; The integration unit should integrate the capabilities of the big data platform into multiple types of tools, forming an MCP service corresponding to each type of tool; The acquisition unit 12 is further configured to acquire the first tool from the first MCP service corresponding to the first intelligent agent.

[0073] Optionally, the apparatus further includes: a parsing unit; The parsing unit is used to use the first intelligent agent to parse the big data development task in combination with historical development tasks to obtain task requirements; use the big model to parse the task requirement information, and select a target tool from the first tool.

[0074] Optionally, the device further includes: a configuration unit; The configuration unit is used to configure response strategies for at least one of the abnormal scenarios, such as agent matching failure, large model access failure, and big data platform call timeout. The response strategies include, but are not limited to, at least one of the following: retry, alarm, backup switchover, and log recording.

[0075] This application provides a task processing device that receives big data development tasks and matches them with a first intelligent agent. It then obtains a first tool corresponding to the first intelligent agent and accesses a large model through a large model adaptation service. Using the large model, it selects a target tool from the first tool for the big data development task and uses the target tool to call a big data platform to execute the big data development task. Therefore, the task processing device proposed in this embodiment adopts a modular design, which can support the expansion of different types of intelligent agents to accommodate big data development tasks with different needs, thus meeting diverse data development requirements. Accessing different large models through the large model adaptation service decouples the large model from the intelligent agent, ensuring that upgrades, updates, and failures of the large model do not affect the performance of the intelligent agent, thereby improving the compatibility of the intelligent agent with the large model.

[0076] Figure 7 This is a schematic diagram of the composition structure of a task processing device 1 provided in an embodiment of this application. In practical applications, based on the same disclosed concept of the above embodiments, such as... Figure 7 As shown, the task processing device 1 in this embodiment includes a processor 15, a memory 16, and a communication bus 17.

[0077] The processor 15 described above can be at least one of the following: Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), CPU, controller, microcontroller, and microprocessor. It is understood that, for different devices, the electronic device used to implement the above processor function can also be other types, and this embodiment does not specifically limit it.

[0078] In this embodiment, the communication bus 17 is used to establish communication between the processor 15 and the memory 16; when the processor 15 executes the running program stored in the memory 16, it implements the following task processing method: Receive a big data development task and match it with a first intelligent agent corresponding to the big data development task; obtain a first tool corresponding to the first intelligent agent; access a big model through a big model adaptation service, and use the big model to select a target tool for the big data development task from the first tool; use the target tool to call the big data platform to execute the big data development task.

[0079] Furthermore, the processor 15 is also configured to receive big data development tasks sent by various clients through various interfaces provided by the intelligent agent routing service, and select the first intelligent agent corresponding to the big data development task.

[0080] Furthermore, the processor 15 is also used to invoke the first model context protocol (MCP) service corresponding to the first intelligent agent using the target tool; and to invoke the big data platform to execute the big data development task using the first MCP service.

[0081] Furthermore, the processor 15 is also used to obtain first result information of the big data platform executing the big data development task; the first result information is machine language result information; access the big model through the big model adaptation service, and use the big model to convert the first result information into second result information; the second result information is natural language result information; and send the second result information.

[0082] Furthermore, the processor 15 is also used to integrate the capabilities of the big data platform into multiple types of tools, forming an MCP service corresponding to each type of tool; and to obtain the first tool from the first MCP service corresponding to the first intelligent agent.

[0083] Furthermore, the processor 15 is also used to utilize the first intelligent agent to analyze the big data development task in conjunction with historical development tasks to obtain task requirements; to analyze the task requirement information using the big model, and to select a target tool from the first tool.

[0084] Furthermore, the processor 15 is also used to configure a response strategy for at least one of the abnormal scenarios, such as agent matching failure, large model access failure, and big data platform call timeout; the response strategy includes, but is not limited to, at least one of the following: retry, alarm, backup switchover, and log recording.

[0085] This application provides a storage medium storing a computer program thereon. The computer-readable storage medium stores one or more programs, which can be executed by one or more processors and applied in a task processing device. The computer program implements the task processing method described above.

[0086] Based on the above embodiments, this application provides a computer program product, including a computer program that can be executed by one or more processors, and the computer program implements the task processing method described above.

[0087] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0088] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the related technology, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause an image display device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this disclosure.

[0089] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application.

Claims

1. A task processing method, characterized in that, The method includes: Receive big data development tasks and match the first intelligent agent corresponding to the big data development tasks; Obtain the first tool corresponding to the first intelligent agent; and access the large model through the large model adaptation service, and use the large model to select the target tool for the big data development task from the first tool; The target tool is used to invoke the big data platform to execute the big data development task.

2. The method according to claim 1, characterized in that, The step of receiving big data development tasks and matching the first intelligent agent corresponding to the big data development tasks includes: The system receives big data development tasks sent by various clients through multiple interfaces provided by the intelligent agent routing service, and matches the first intelligent agent corresponding to the big data development task.

3. The method according to claim 1, characterized in that, The step of using the target tool to call the big data platform to execute the big data development task includes: The target tool is used to invoke the first model context protocol (MCP) service corresponding to the first intelligent agent. The first MCP service is used to call the big data platform to execute the big data development task.

4. The method according to claim 1, characterized in that, After the method involves using the target tool to invoke the big data platform to execute the big data development task, the method further includes: Obtain the first result information of the big data platform executing the big data development task; the first result information is machine language result information. The large model is accessed through the large model adaptation service, and the first result information is converted into second result information using the large model; the second result information is natural language result information. Send the second result information.

5. The method according to claim 1, characterized in that, Before obtaining the first tool corresponding to the first intelligent agent, the method further includes: The capabilities of the big data platform are integrated into multiple types of tools, forming an MCP service corresponding to each type of tool; The first tool for obtaining the first intelligent agent includes: The first tool is obtained from the first MCP service corresponding to the first intelligent agent.

6. The method according to claim 1, characterized in that, The step of selecting a target tool for the big data development task from the first tool using the big model includes: Using the first intelligent agent, the big data development task is analyzed in conjunction with historical development tasks to obtain the task requirements; The task requirement information is analyzed using the large model, and a target tool is selected from the first tool.

7. The method according to claim 1, characterized in that, The method further includes: Configure response strategies for at least one of the following abnormal scenarios: agent matching failure, large model access failure, and big data platform call timeout; the response strategies include, but are not limited to, at least one of the following: retry, alarm, backup switchover, and log recording.

8. A task processing device, characterized in that, The device includes: a processor, a memory, and a communication bus; when the processor executes the running program stored in the memory, it implements the method as described in any one of claims 1-7.

9. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, The computer program, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images