Intelligent agent arrangement method and system based on large model

Abstract

The invention relates to the technical field of model development, and discloses an agent arrangement method and system based on a large model, and the method comprises the following steps: receiving and processing information from a project management tool and a technology source, extracting key content, arranging the key content into a structured change request list, and storing the change request list in a database; based on the structured change request list and the existing configuration information, the specific influence of each change request on the CI / CD process is predicted, and a group of reasonable configuration modification schemes is output according to the prediction result. According to the method, each link in the CI / CD process can be automatically processed, continuous optimization can be realized through a data driving mode, meanwhile, the change of requirements and technologies can be quickly responded, manual configuration change is minimized, introduction of human errors is avoided, CI / CD streams are continuously optimized through data analysis and feedback circulation, self-adjustment and optimization are carried out by utilizing an AI intelligent agent, and the CI / CD flow optimization efficiency is improved. And the dependence on human operators is reduced.

Description

Technical Field

[0001] The present invention relates to the field of model development, and more specifically, it relates to an intelligent agent orchestration method and system based on large models. Background Art

[0002] With the development of artificial intelligence (AI) technology, especially the application of large models, it has become particularly important to build a CI / CD pipeline that can automatically integrate new AI modules and conduct comprehensive performance evaluations.

[0003] During the application of large models, the rapidly changing requirements and technical environment require frequent adjustments to the CI / CD pipeline configuration, which may increase complexity and introduce the risk of human error. Summary of the Invention

[0004] The present invention provides an intelligent agent orchestration method and system based on large models to solve the technical problems in the related art.

[0005] The present invention provides an intelligent agent orchestration method based on large models, including the following steps:

[0006] S100: Requirement and technology change monitoring, by receiving and processing information from project management tools and technical sources, extracting key content, and organizing it into a structured list of change requests;

[0007] S200: Change impact assessment and planning, based on the structured list of change requests and existing configuration information, predicting the specific impact of each change request on the CI / CD process, and according to the prediction results, outputting a set of reasonable configuration modification plans, and selecting the most suitable plan as the final implementation plan through cost-benefit analysis;

[0008] S300: Automated configuration adjustment, based on the final implementation plan of S200, testing the new configuration in a non-production environment, and verifying the new configuration through a detailed deployment instruction set, and outputting a verification result showing that the new configuration passes all inspections and is ready to be promoted to the production environment;

[0009] S400: Continuous monitoring and feedback, based on the new configuration shown in the verification result, real-time tracking the system performance under the new configuration, and once an abnormal situation is detected, immediately triggering a rollback mechanism to restore to the previous stable state, collecting the system operation status, and outputting a health status report and abnormal event records;

[0010] S500: Data-driven continuous optimization, according to the health status report and abnormal event records, generating an optimization strategy through an optimization strategy generation model, and continuously optimizing the CI / CD process in S100 - S400 according to the optimization strategy.

[0011] Furthermore, S100 also includes the following steps:

[0012] S110, Receive project management tool information and technical information;

[0013] S120, Use natural language processing technology to parse messages from project management tools and extract key information;

[0014] S130, Monitor technical information and extract release information of the latest frameworks, libraries or tools;

[0015] S140, Organize the collected key information and release information into a structured list of change requests.

[0016] Furthermore, S200 also includes the following steps:

[0017] S210, Predict the impact scope of each change request based on the structured list of change requests in S100 and the existing configuration information;

[0018] S220, Automatically generate a set of recommended configuration modification solutions based on the prediction results;

[0019] S230, Compare the cost - effectiveness of multiple solutions and select the optimal solution as the final implementation plan.

[0020] Furthermore, S300 also includes the following steps:

[0021] S310, Input the final implementation plan in S200, test the sandbox environment of the new configuration, and prove the new configuration settings that are feasible after non - production environment testing;

[0022] S320, Deploy the instruction set. The instruction set refers to the detailed steps for applying the new configuration to the production environment.

[0023] Furthermore, the specific steps for applying to the production environment in S320 are as follows:

[0024] S321, Generate detailed configuration change instructions: Through the final implementation plan, which includes introducing new continuous integration tools and updating existing test scripts, output detailed configuration change instructions;

[0025] S322, Use declarative configuration management tools: Configuration change instructions, Helm Charts configuration file paths, Kubernetes cluster API keys, and output that the new tools and updated test scripts are successfully deployed to the non - production Kubernetes cluster;

[0026] S323, Verify the change effect: Based on the system performance metrics and user feedback under the new configuration, output the verification result showing that the new configuration passes all checks and is ready to be promoted to the production environment.

[0027] Furthermore, S400 also includes the following steps:

[0028] S410, input the new configuration and real-time monitoring data stream corresponding to the verification result in S300; The real-time monitoring data stream includes information such as server load, error logs, and user behavior from the monitoring system;

[0029] Among them, the real-time monitoring data stream includes information such as server load, error logs, and user behavior from the monitoring system;

[0030] S420, real-time track the performance of the deployed system, mainly the part directly related to the change. If an abnormal situation is detected, immediately trigger the rollback mechanism to restore to the previous stable state;

[0031] S430, collect the operation status of the system under the new configuration, list all detected problems and their handling situations, and output a health status report and abnormal event records.

[0032] Furthermore, the following content is also included in S200:

[0033] The calculation formula for change impact assessment is as follows:

[0034] ;

[0035] Among them, refers to the specific impact scope of the predicted change on the existing system, is the first input feature vector, containing detailed information about requirements and technical changes, are model parameters, where represents the transpose of the weight matrix, represents the weight matrix for change impact assessment, represents the bias term for change impact assessment;

[0036] ;

[0037] Among them represents the requirement change request, represents the technical change information, represents the context information;

[0038] The calculation formula for generating the recommended configuration modification plan is as follows:

[0039] ;

[0040] Among them, represents a set of possible configuration modification plans and their expected effects, represents the decision logic of a decision tree model, is a historical dataset that contains successful cases of past similar changes and corresponding best practices, represents the decision path constructed based on the input features and historical data, is the feature extraction function based on the first input feature vector ;

[0041] ;

[0042] where represents the change records that have been successfully implemented, represents the best practice guidelines;

[0043] ;

[0044] where, represents the objective function, is the second input feature vector that contains detailed information about each recommended configuration modification plan, is the cost function, is the benefit function, and are the weights of cost and benefit respectively;

[0045] ;

[0046] where represents the specific configuration modification content, represents the context information;

[0047] ;

[0048] where represents the implementation time cost of the i-th plan, including the time required for development, testing, and deployment, represents the resource consumption cost of the i-th plan, represents the risk cost of the i-th plan;

[0049] where represents the performance improvement of the i-th plan, represents the quality improvement of the i-th plan, represents the user experience improvement of the i-th plan;

[0050] The calculation formula for optimal solution selection is as follows:

[0051] ;

[0052] where represents the best solution, that is, the optimal solution.

[0053] Further, the calculation formula of the optimization strategy generation model in S500 is as follows:

[0054] ;

[0055] where represents the optimization strategy, which is an improvement suggestion for the future CI / CD process, represents the historical data set, including past change records and corresponding performance changes, represents the current feedback, including the health status report and abnormal event records, are the model parameters of the optimization strategy generation model, represents the reinforcement learning function;

[0056] ;

[0057] where represents the health status report, represents the abnormal event record.

[0058] The present invention also proposes an intelligent agent orchestration system based on a large model, which executes one or more steps in the foregoing intelligent agent orchestration method based on a large model, including:

[0059] Information collection and parsing module: By receiving and processing information from project management tools and technical sources, extracting key content, and organizing it into a structured list of change requests;

[0060] Change impact assessment and planning module: Based on the structured list of change requests and existing configuration information, predicting the specific impact of each change request on the CI / CD process, and according to the prediction results, outputting a set of reasonable configuration modification plans, and selecting the most suitable plan as the final implementation plan through cost-benefit analysis;

[0061] Automated configuration adjustment module: Based on the final implementation plan, testing the new configuration in a non-production environment, and verifying the new configuration through a detailed deployment instruction set, and outputting a verification result indicating that the new configuration passes all checks and is ready to be promoted to the production environment;

[0062] Continuous monitoring and feedback module: Based on the new configuration shown in the verification result, real-time tracking of the system performance under the new configuration, immediately triggering a rollback mechanism once an abnormal situation is detected, restoring to the previous stable state, collecting the system operation status, and outputting a health status report and abnormal event records;

[0063] Data-driven continuous optimization module: According to the health status report and abnormal event records, outputting an optimization strategy through the optimization strategy generation model, and continuously optimizing the CI / CD process according to the optimization strategy.

[0064] The present invention also provides a storage medium storing non - transitory computer - readable instructions for performing one or more steps in the aforementioned sustainable communication optimization processing method in a restricted connection environment.

[0065] The beneficial effects of the present invention are as follows:

[0066] The present invention can automate each link in the CI / CD process, continuously optimize through a data - driven approach, quickly respond to changes in requirements and technologies, minimize manual configuration changes, avoid introducing human errors, continuously optimize the CI / CD flow through data analysis and feedback loops, use AI agents for self - adjustment and optimization, and reduce dependence on human operators. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 is a flowchart of an intelligent agent orchestration method based on a large - model proposed by the present invention;

[0068] Figure 2 is a structural block diagram of an intelligent agent orchestration system based on a large - model proposed by the present invention;

[0069] Figure 3 is a process diagram of an intelligent agent orchestration method based on a large - model proposed by the present invention.

[0070] In the figure: 101, information collection and parsing module; 102, change impact assessment and planning module; 103, automated configuration adjustment module; 104, continuous monitoring and feedback module; 105, data - driven continuous optimization module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0071] Now, the subject matter described herein will be discussed with reference to exemplary embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein. Without departing from the scope of protection of the content of this specification, the functions and arrangements of the elements discussed can be changed. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described relative to some examples can also be combined in other examples.

[0072] As Figure 1 shown, an intelligent agent orchestration method based on a large - model includes the following steps:

[0073] S100: Monitoring requirements and technology changes, by receiving and processing information from project management tools and technology sources, extracting key content, and organizing it into a structured list of change requests;

[0074] S110, receiving project management tool information and technology information;

[0075] Among them, the project management tool information includes task updates and issue reports of project management tools (such as Jira, Trello, etc.);

[0076] Among them, the technical information includes new version release notes, best practice guides, etc. from technical blogs, community forums, and official documents;

[0077] S120, use natural language processing (NLP) technology to parse messages from project management tools and extract key information;

[0078] S130, monitor technical information and extract release information of the latest frameworks, libraries, or tools;

[0079] S140, organize the collected key information and release information into a structured change request list;

[0080] Among them, the structured change request list contains the identified requirements and technical changes and their priorities.

[0081] S200: Change impact assessment and planning. Based on the structured change request list and existing configuration information, predict the specific impact of each change request on the CI / CD process. According to the prediction results, output a set of reasonable configuration modification plans, and through cost-benefit analysis, select the most suitable plan as the final implementation plan;

[0082] In an embodiment of the present invention, the specific steps are as follows:

[0083] S210, predict the impact scope of each change request according to the structured change request list in S100 and the existing configuration information;

[0084] Among them, the existing configuration information includes existing CI / CD configurations, the status of running services, etc.;

[0085] In an embodiment of the present invention, the calculation formula for change impact assessment is as follows:

[0086] ;

[0087] Among them, is the first input feature vector, containing detailed information about requirements and technical changes, are model parameters, including weights and biases, etc., where represents the transpose of the weight matrix, represents the weight matrix for change impact assessment, used to measure the importance of different input features, represents the bias term for change impact assessment, adjusting the basic level of the prediction result;

[0088] ;

[0089] Among them represents a requirements change request (such as adding new functions, modifying existing features, etc.), represents technical change information (such as the introduction of a new framework, library, or tool), represents context information (such as the current CI / CD configuration, the status of running services, etc.);

[0090] Among them (Change Impact) refers to the specific scope of the predicted impact of the change on the existing system, which is a classification label (such as "low impact", "medium impact", "high impact") or a continuous value (such as the percentage of performance degradation);

[0091] S220, according to the prediction result, automatically generate a set of recommended configuration modification plans;

[0092] In an embodiment of the present invention, the calculation formula for generating the recommended configuration modification plan is as follows:

[0093] ;

[0094] Among them, (Recommended Configurations) represents a set of possible configuration modification plans and their expected effects, and the possibility of each plan being adopted can be represented by a probability distribution, represents the decision logic of a decision tree model, is a historical data set, which contains successful cases of past similar changes and corresponding best practices, represents the decision path constructed based on the input features and historical data, is based on the input feature vector of the feature extraction function;

[0095] ;

[0096] Among them represents the change records successfully implemented, represents the best practice guidelines.

[0097] S230, compare the cost-benefit of multiple plans, and select the optimal solution as the final implementation plan;

[0098] The calculation formula for comparing the cost-benefit of multiple plans is as follows:

[0099] ;

[0100] Among them, (Objective Function) represents the objective function, is the second input feature vector, containing detailed information about each recommended configuration modification plan, is the cost function, measuring the resource consumption required to implement the plan, is the benefit function, measuring the improvements or benefits brought about after implementing the plan, and are the weights of cost and benefit respectively, reflecting their relative importance in decision-making.

[0101] ;

[0102] where represents the specific configuration modification content, represents the context information (such as the current CI / CD configuration, the status of running services, etc.);

[0103] ;

[0104] where represents the implementation time cost of the i-th plan, including the time required for development, testing, and deployment, represents the resource consumption cost of the i-th plan, such as server resources, license fees, etc., represents the risk cost of the i-th plan, evaluating the potential risks and the costs of their countermeasures.

[0105] where represents the performance improvement of the i-th plan, such as shortened response time, increased throughput, etc., represents the quality improvement of the i-th plan, such as reduced error rate, improved stability, etc., represents the improvement of the user experience of the i-th plan, such as an increase in the user satisfaction score, etc.

[0106] The calculation formula for optimal solution selection is as follows:

[0107] ;

[0108] where (Optimal Solution) represents the best solution, that is, the optimal solution. If the benefit is a positive contribution, a negative sign is used in the objective function;

[0109] S300: Automated configuration adjustment. Based on the final implementation plan of S200, the new configuration is tested in the non-production environment, and the new configuration is verified through a detailed deployment instruction set. The output verification result shows that the new configuration passes all checks and is ready to be promoted to the production environment;

[0110] In one embodiment of the present invention, it specifically includes the following steps:

[0111] S310, Input the final implementation plan in S200, test the newly configured sandbox environment, and the newly configured settings that are proven feasible after non-production environment testing;

[0112] In one embodiment of the present invention, the calculation formula for testing the newly configured sandbox environment is as follows:

[0113] ;

[0114] Where (Old Config) is the configuration in the existing production environment, including existing CI / CD configurations, service status, etc., (Tested Config) represents the newly configured settings verified in the non-production environment, represents the process of applying a certain change or configuration to the existing system;

[0115] S320, Deploy the instruction set. The instruction set refers to the detailed steps of applying the new configuration to the production environment;

[0116] In one embodiment of the present invention, the specific steps for applying to the production environment are as follows:

[0117] S321, Generate detailed configuration change instructions: Through the final implementation plan, including introducing new continuous integration tools and updating existing test scripts, output detailed configuration change instructions, such as commands for installing new tools, specific locations for updating test scripts, etc.;

[0118] S322, Use a declarative configuration management tool: Configure the configuration change instructions, Helm Charts configuration file path, Kubernetes cluster API key, and output that the new tools and updated test scripts are successfully deployed to the non-production Kubernetes cluster;

[0119] S323, Verify the change effect: According to the system performance metrics under the new configuration (such as a 15% reduction in build time, a 10% increase in test coverage), user feedback (such as no new error reports), output the verification result showing that the new configuration passes all checks and is ready to be promoted to the production environment;

[0120] The calculation formula for verifying the change effect is as follows:

[0121] ;

[0122] Where represents the monitoring metrics, including system performance metrics (such as CPU utilization, memory usage, etc.), user feedback (such as response time, error reports, etc.), Represents the verification model parameters, including threshold setting (such as the maximum allowable percentage of performance degradation), anomaly detection algorithms (such as statistical-based anomaly detection, machine learning models), where (ValidationResult) represents the verification result, including whether the verification is passed, the problems found and their severity levels, Represents the verification function.

[0123] S400: Continuous monitoring and feedback. Based on the new configuration shown in the verification result, continuously track the system performance under the new configuration in real time. Once an abnormal situation is detected, immediately trigger the rollback mechanism to restore to the previous stable state, collect the system health status, and output the health status report and abnormal event records;

[0124] In an embodiment of the present invention, it specifically includes the following steps:

[0125] S410, input the new configuration corresponding to the verification result in S300 and the real-time monitoring data stream;

[0126] where the real-time monitoring data stream includes information such as server load, error logs, and user behavior from the monitoring system;

[0127] S420, continuously track the system performance after deployment, mainly the part directly related to the change. If an abnormal situation is detected, immediately trigger the rollback mechanism to restore to the previous stable state;

[0128] S430, collect the system operation status under the new configuration, list all detected problems and their handling situations, and output the health status report and abnormal event records;

[0129] S500: Data-driven continuous optimization. According to the health status report and abnormal event records, generate an optimization strategy through the optimization strategy generation model, and continuously optimize the CI / CD process in S100 - S400 according to the optimization strategy;

[0130] where the optimization strategy refers to the improvement suggestions for the CI / CD process.

[0131] In an embodiment of the present invention, the calculation formula of the optimization strategy generation model is as follows:

[0132] ;

[0133] where (Optimization Strategy) represents the optimization strategy, which is the improvement suggestions for the future CI / CD process, and represents the specific optimization measures with a rule set or action sequence, Represents the historical dataset, including past change records and corresponding performance changes, Represents the current feedback, including the system feedback after the most recent change (such as health status reports and abnormal event records), Are model parameters, including weights and biases, etc., Represents the reinforcement learning function.

[0134] ;

[0135] Among them Represents the health status report, Represents the abnormal event record.

[0136] The steps to optimize the training of the policy generation model include:

[0137] Step 1, construct a target network, and the target network replicates the optimization policy generation model;

[0138] Step 2, randomly obtain an experience from the experience pool, and the experience includes the state at time e , the action executed at time e , the state at time e+1 , the reward obtained by executing the action ; ;

[0139] The states at times e and e+1 respectively represent the current feedback information of the e-th and e+1-th moments input to the optimization policy generation model;

[0140] Then calculate the probability value corresponding to output by the target network under the state ;

[0141] ;

[0142] Represents the maximum probability value output by the target network under the state;

[0143] Step 3, according to and calculate the loss value loss and update the optimization policy generation model;

[0144] ;

[0145] Represents the probability value corresponding to output by the optimization policy generation model under the state;

[0146] is the discount factor, a value between 0 and 1;

[0147] Step 4: Update the target network every fixed number of network updates so that its parameters are the same as those of the current optimized policy generation model;

[0148] Step 5: Terminate the steps until the number of collected experiences reaches the set threshold or the optimized policy generation model converges.

[0149] In one embodiment of the present invention, the experience is derived from the records of historical action executions.

[0150] In one embodiment of the present invention, the reward has the following calculation formula:

[0151] ;

[0152] where represents the current state at time e, which is jointly determined by historical data and the current feedback and represents the action taken at time e, that is, the specific optimization strategy.

[0153] where represents the performance improvement, ;

[0154] where represents the new performance metric after implementing the optimization strategy (such as build time, test coverage, etc.), represents the old performance metric before implementation.

[0155] where represents the cost savings, ;

[0156] where represents the resource consumption cost after implementing the optimization strategy (such as server resources, license fees, etc.), represents the resource consumption cost before implementation.

[0157] where represents the reduction in failure rate, ;

[0158] where represents the number of failure rates or abnormal events after implementing the optimization strategy, represents the number of failure rates or abnormal events before implementation.

[0159] where represents the time savings, ;

[0160] where represents the total deployment time after implementing the optimization strategy, and represents the total deployment time before implementation.

[0161] Among them , , and respectively represent the weights corresponding to each parameter, reflecting their relative importance in the decision-making.

[0162] At least one embodiment of the present disclosure provides an agent orchestration system based on a large model, which executes one or more steps in the foregoing method for agent orchestration based on a large model, including:

[0163] Information collection and parsing module 101: By receiving and processing information from project management tools and technical sources, extracting key content, and organizing it into a structured list of change requests;

[0164] Change impact assessment and planning module 102: Based on the structured list of change requests and existing configuration information, predicting the specific impact of each change request on the CI / CD process, and according to the prediction results, outputting a set of reasonable configuration modification plans, and selecting the most suitable plan as the final implementation plan through cost-benefit analysis;

[0165] Automated configuration adjustment module 103: Based on the final implementation plan, testing the new configuration in a non-production environment, and verifying the new configuration through a detailed deployment instruction set, and outputting a verification result indicating that the new configuration passes all checks and is ready to be promoted to the production environment;

[0166] Continuous monitoring and feedback module 104: Based on the new configuration shown in the verification result, real-time tracking of the system performance under the new configuration, and once an abnormal situation is detected, immediately triggering a rollback mechanism to restore to the previous stable state, collecting the system operation status, and outputting a health status report and abnormal event records;

[0167] Data-driven continuous optimization module 105: According to the health status report and abnormal event records, generating an optimization strategy through an optimization strategy generation model, and continuously optimizing the CI / CD process according to the optimization strategy.

[0168] At least one embodiment of the present disclosure provides a storage medium storing non-temporary computer-readable instructions for executing one or more steps in the foregoing method for sustainable communication optimization processing in a restricted connection environment.

[0169] A computer program may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.

[0170] The above has described the embodiments of this example, but this example is not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative rather than restrictive. Under the inspiration of this example, those of ordinary skill in the art can also make many forms, all of which fall within the protection scope of this example.

Claims

1. An intelligent agent orchestration method based on a large model, characterized in that It includes the following steps: S100: Requirement and technology change monitoring. By receiving and processing information from project management tools and technology sources, extracting key content, and organizing it into a structured list of change requests; S200: Change impact assessment and planning. Based on the structured list of change requests and existing configuration information, predict the specific impact of each change request on the CI / CD process. According to the prediction results, output a set of reasonable configuration modification plans, and through cost-benefit analysis, select the most suitable plan as the final implementation plan; S300: Automated configuration adjustment. Based on the final implementation plan of S200, test the new configuration in a non-production environment and verify the new configuration through a detailed deployment instruction set. The output verification result shows that the new configuration passes all checks and is ready to be promoted to the production environment; S400: Continuous monitoring and feedback. Based on the new configuration shown in the verification result, real-time track the system performance under the new configuration. Once an abnormal situation is detected, immediately trigger the rollback mechanism to restore to the previous stable state, collect the system operation status, and output a health status report and abnormal event records; S5 00: Data-driven continuous optimization. According to the health status report and abnormal event records, generate optimization strategies through an optimization strategy generation model, and continuously optimize the CI / CD process in S100 - S400 according to the optimization strategies.

2. The intelligent agent orchestration method based on a large model according to claim 1, wherein S100 also includes the following steps: S110, Receive project management tool information and technical information; S120, Use natural language processing technology to parse messages from project management tools and extract key information; S130, Monitor technical information and extract release information of the latest frameworks, libraries, or tools; S140, Organize the collected key information and release information into a structured list of change requests.

3. The intelligent agent orchestration method based on a large model according to claim 2, characterized in that, S200 also includes the following steps: S210, Predict the impact scope of each change request according to the structured list of change requests in S100 and the existing configuration information; S220, Automatically generate a set of recommended configuration modification plans according to the prediction results; S230, Compare the cost-benefit of multiple plans and select the optimal solution as the final implementation plan.

4. An agent orchestration method based on a large model according to claim 3, characterized in that, S300 also includes the following steps: S310, Input the final implementation plan of S200, test the sandbox environment of the new configuration, and set the new configuration that has been proven feasible after non-production environment testing; S320, Deployment instruction set. The instruction set refers to the detailed steps of applying the new configuration to the production environment.

5. The intelligent agent orchestration method based on a large model according to claim 4, wherein, The specific steps for applying to the production environment in S320 are as follows: S321, Generate detailed configuration change instructions: Through the final implementation plan, which includes introducing new continuous integration tools and updating existing test scripts, output detailed configuration change instructions; S322, Use declarative configuration management tools: Configuration change instructions, Helm Charts configuration file paths, Kubernetes cluster API keys, and output that the new tools and updated test scripts are successfully deployed to the non-production Kubernetes cluster; S323. Verify the change effect: Based on the system performance metrics and user feedback under the new configuration, output the verification result showing that the new configuration passes all checks and is ready to be promoted to the production environment.

6. The intelligent agent orchestration method based on a large model according to claim 5, wherein, S400 further includes the following steps: S410, Input the verification result in S300 The corresponding new configuration and real-time monitored data stream; Among them, the real-time monitored data stream includes information such as server load, error logs, and user behavior from the monitoring system; S420. Real-time track the performance of the deployed system, mainly the parts directly related to the change. If an abnormal situation is detected, immediately trigger the rollback mechanism to restore to the previous stable state; S430. Collect the operating conditions of the system under the new configuration, list all detected problems and their handling situations, and output a health status report and an abnormal event record.

7. An agent orchestration method based on a large model according to claim 6, wherein, The following content is also included in S200: Among them, the calculation formula for change impact assessment is as follows: ； Among them, refers to the specific impact scope of the predicted change on the existing system, is the first input feature vector, containing detailed information about requirements and technical changes, are model parameters, where represents the transpose of the weight matrix, represents the weight matrix for change impact assessment, represents the bias term for change impact assessment; ； Among them represents a requirements change request, represents technical change information, represents context information; Among them, the calculation formula for generating the recommended configuration modification plan is as follows: ； Among them, represents a set of possible configuration modification scenarios and their expected effects, represents the decision logic of a decision tree model, is a historical data set containing successful cases of past similar changes and corresponding best practices, represents a decision path constructed based on input features and historical data, is based on the first input feature vector feature extraction function; ； Among them represents the change record successfully implemented, represents the best practice guide; ； wherein, represents the objective function, is the second input feature vector, which contains detailed information about each recommended configuration modification plan, is the cost function, is the revenue function, and are the weights of cost and revenue respectively; ； Among them represents the specific configuration modification content, represents the context information; ； ； Among them represents the implementation time cost of the i-th solution, including the time required for development, testing, and deployment, represents the resource consumption cost of the i-th solution, represents the risk cost of the i-th solution; Among them represents the performance improvement of the i-th solution, represents the quality improvement of the i-th solution, represents the improvement of the user experience of the i-th solution; The calculation formula for optimal solution selection is as follows: ； Among them represents the best solution, that is, the optimal solution.

8. An agent orchestration method based on a large model according to claim 7, characterized in that, The calculation formula of the optimization strategy generation model in S500 is as follows: ； Among them represents an optimization strategy, which is an improvement suggestion for the future CI / CD process represents a historical dataset, including past change records and corresponding performance changes represents the current feedback, including health status reports and abnormal event records are the model parameters of the optimization strategy generation model represents a reinforcement learning function ； Among them represents the health status report, represents the abnormal event record.

9. An intelligent agent orchestration system based on a large model, which executes one or more steps in the method for orchestrating intelligent agents based on a large model according to any one of claims 1-8, characterized in that, Including: Information collection and parsing module: By receiving and processing information from project management tools and technical sources, extract key content and organize it into a structured list of change requests; Change impact assessment and planning module: Based on the structured list of change requests and existing configuration information, predict the specific impact of each change request on the CI / CD process. According to the prediction results, output a set of reasonable configuration modification plans, and select the most suitable plan as the final implementation plan through cost-benefit analysis; Automated configuration adjustment module: Based on the final implementation plan, test the new configuration in the non-production environment and verify the new configuration through a detailed deployment instruction set. Output the verification result showing that the new configuration passes all checks and is ready to be promoted to the production environment; Continuous monitoring and feedback module: Based on the new configuration shown in the verification result, real-time track the performance of the system under the new configuration. Once an abnormal situation is detected, immediately trigger the rollback mechanism to restore to the previous stable state, collect the system operating conditions, and output a health status report and an abnormal event record; Data-driven continuous optimization module: According to the health status report and abnormal event record, output an optimization strategy through the optimization strategy generation model, and continuously optimize the CI / CD process according to the optimization strategy.

10. A storage medium, characterized in that, Store non-temporary computer-readable instructions for executing one or more steps in the sustainable communication optimization processing method in any one of claims 1-8 in a restricted connection environment.

Images