Cross-end modular tool execution method, system, electronic device, and storage medium

Abstract

This invention discloses a cross-platform modular tool execution method, system, electronic device, and storage medium. The method includes the following steps: receiving a tool invocation request initiated by a user on a client; obtaining the preset attribute tags of the invoked tool and routing it to a local edge execution path or a cloud collaboration path; the local edge execution path includes: initializing a web page assembly technology module and a web page worker thread on the client; performing local calculations on the client and outputting the results; the cloud collaboration path includes: concatenating execution instructions on the server side, sending the concatenated instructions to an AI model to generate structured results, performing pattern consistency verification on the structured results; and sending the verified results to the client's web page worker thread for isolated execution. This invention, through a local priority strategy and cloud-based artificial intelligence collaboration, keeps sensitive tasks processed on the user's device, not only improving privacy and security but also achieving intelligent boundary expansion.

Description

Technical Field

[0001] This invention relates to the fields of Internet application architecture and artificial intelligence technology, and in particular to cross-platform modular tool execution methods, systems, electronic devices and storage media. Background Technology

[0002] Traditional online tools (such as OCR text recognition, regular expression generation, and mock data simulation) typically employ a pure front-end or pure back-end architecture, which presents the following technical problems: Data privacy breach risks: Pure backend solutions force users to upload sensitive data (such as ID photos and internal company code) to remote servers, increasing the risk of data hijacking or illegal misuse during transmission and storage. Browser performance degradation: Pure frontend solutions typically run in a single-threaded environment when handling high-load computations (such as loading large libraries or complex transformation logic), which can easily lead to blocking of the main page thread, causing UI lag or even browser crashes. Uneven allocation of computing resources: Current technologies lack a mechanism to flexibly allocate resources based on task characteristics (such as whether complex semantic understanding is involved), resulting in simple tasks consuming cloud computing power or complex tasks failing to execute locally. High cost of AI integration: Directly calling large models from the frontend often faces problems such as unstable response formats, leaked prompts, and redundant request logic, lacking an efficient semantic mapping and transformation layer.

[0003] In summary, there is a need to design a cross-platform modular tool execution method, system, electronic device, and storage medium to solve the aforementioned problems in the prior art. Summary of the Invention

[0004] This invention provides a cross-platform modular tool execution method, system, electronic device, and storage medium, which solves the problem of the single architecture of online tools in the prior art.

[0005] To achieve the goal of solving the above-mentioned technical problems, the present invention adopts the following technical solution: The cross-platform modular tool execution method includes the following steps: Receive tool invocation requests initiated by users on the client; Obtain the preset attribute tags of the invoked tool, and route the invocation request to the local edge execution path or the cloud collaboration path according to the preset attribute tags; The local edge execution path includes: initializing the web page assembly technology module and / or web page worker thread on the client; performing local computation using the client's local computing resources and outputting the results; The cloud-based collaborative path includes: the server executes instruction concatenation, sends the concatenated instructions to the AI ​​model to generate a structured result, performs pattern consistency verification on the structured result, and sends the verified result to the client's web page worker thread for isolated execution.

[0006] In some embodiments of the present invention, the server-side execution instruction splicing includes: structurally splicing together a preset instruction template, tool-specific mode constraint specifications, user input context environment and original natural language description to forcibly constrain the output boundary of the AI ​​model.

[0007] In some embodiments of the present invention, the generation of structured results by the AI ​​model includes: determining a primary AI model based on the context length and logical complexity of the call request; executing instructions using the primary AI model; capturing the abnormal state if the primary AI model fails to execute, fails to meet preset performance indicators, or triggers a content security policy; and then switching to a backup AI model, which continues to execute the concatenated instructions based on the abnormal state; if all backup AI models fail to recover from the disaster, a "safe rollback" instruction is executed, i.e., a structured error object is output to the client; otherwise, the backup AI model outputs the structured result.

[0008] In some embodiments of the present invention, the pattern consistency verification includes: after performing noise filtering and format completion on the structured result; extracting the key structure in the structured result; determining whether the key structure conforms to the tool-specific pattern constraint specification, and if so, outputting the structured result; otherwise, triggering a repair loop.

[0009] In some embodiments of the present invention, the repair loop includes: outputting an error log when the critical structure does not conform to the tool-specific pattern constraint specification; inputting the error log, the original user instruction, and the tool-specific pattern constraint specification into a backup AI model; the backup AI model executing the instruction and outputting a structured result; and continuing to perform pattern consistency verification on the structured result output by the backup AI model.

[0010] In some embodiments of the present invention, the isolated execution of the web page worker thread sent to the client includes: mounting a global exception listener and a timeout circuit breaker in the client sandbox to intercept exponential operations or infinite loops caused by uncontrollable divergent logic, thereby avoiding lag or crashes in the main page rendering process.

[0011] In some embodiments of the present invention, after receiving a tool call request initiated by a user on the client, the method further includes reading the user's mode selection, wherein the mode includes a standard mode and an artificial intelligence mode; the standard mode is the local edge execution path, and the artificial intelligence mode is the cloud collaboration path.

[0012] In some embodiments of the present invention, a cross-platform modular tool execution system, for implementing the above-described cross-platform modular tool execution method, includes: The request reading module is used to read the user's tool call requests; The path selection module is used to determine the execution path based on user selection or the tool's preset attribute labels; The request processing module is used to select an AI model based on the user's request and output structured results. An isolated execution module is used to execute the structured results on the client side.

[0013] In some embodiments of the present invention, an electronic device is provided, comprising: A processor, and a memory and a transceiver communicatively connected to the processor; The memory stores computer-executed instructions; the transceiver is used for sending and receiving data. The processor executes computer execution instructions stored in the memory to implement the cross-end modular tool execution method described above.

[0014] In some embodiments of the present invention, a computer-readable storage medium is provided, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the aforementioned cross-end modular tool execution method.

[0015] The technical solution of the present invention has the following technical effects compared with the prior art: This invention employs a "local priority" strategy to keep sensitive tasks processed on the user device, achieving "data not leaving the domain" and enhancing privacy and security. It utilizes multithreading and binary instruction sets for acceleration, addressing the impact of complex tasks on user interface responsiveness and optimizing performance. Through cloud-based artificial intelligence collaboration, it enables traditional static tools to understand natural language and process structured semantics, achieving intelligent boundary expansion. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the cross-platform modular tool execution method shown in Example 1.

[0018] Figure 2 This is a schematic diagram of the cross-end modular tool execution system shown in Example 2.

[0019] Figure 3 This is a schematic diagram of the structure of the electronic device.

[0020] Reference numerals: 100, Cross-platform modular tool execution system; 110, Request reading module; 120, Path selection module; 130, Request processing module; 140, Isolated execution module; 200, Electronic device; 210, Processor; 220, Memory; 230, Transceiver. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0023] Example 1, referring to Figure 1 As shown, the cross-platform modular tool execution method includes the following steps: Step S1: Receive the tool call request initiated by the user on the client; Step S2: Obtain the preset attribute tags of the invoked tool, and route the call request to the local edge execution path or the cloud collaboration path according to the preset attribute tags.

[0024] S21. When a user-initiated request task is received, the characteristics of the request task are identified to determine the task type, and a preliminary classification is performed based on the task type.

[0025] For example, when the feature of the requested task is identified as "Base encoding", the task type is determined to be a basic transformation; when the feature of the requested task is identified as "natural language generation regular expression", the task type is determined to be a complex semantic task.

[0026] S22. Each tool has preset attribute tags, and the traffic is automatically distributed according to its attribute tags.

[0027] In some embodiments, when a user invokes a tool that is marked as "highly privacy sensitive" (such as local OCR document recognition) or "purely deterministic computation" (such as Base64 encoding), it is automatically routed to the local edge path by default; when the invoked tool relies on strong semantic capabilities such as natural language generation and fuzzy instruction parsing, it is automatically routed to the cloud collaborative path.

[0028] Meanwhile, in some embodiments of this application, a manual intervention switch between "standard mode" and "artificial intelligence mode" is retained in the user interface to deal with special scenarios.

[0029] In summary, when the user selects "Standard Mode," the local edge execution path includes: initializing the web page assembly technology module and / or web page worker threads on the client side; performing local computation on the client side using local computing resources (including but not limited to binary instruction sets) and outputting the results; When the user selects "Artificial Intelligence Mode", the cloud collaboration path includes: the server executes instruction concatenation, sends the concatenated instructions to the AI ​​model to generate a structured result, performs pattern consistency verification on the structured result, and sends the verified result to the client's web page worker thread for isolated execution.

[0030] The execution steps of the cloud collaboration path will be described in detail in step S3.

[0031] S31. Server-side instruction concatenation: Structured concatenation of preset instruction templates, tool-specific mode constraints, user input context, and original natural language descriptions to enforce constraints on the output boundaries of the AI ​​model.

[0032] For example, the server sets hard constraint rules for each tool. For instance, for the "regular expression generator", the constraint logic is: "According to the JSON Schema specification, the output must be valid JSON and contain pattern and explanation fields."

[0033] S32. Determine the primary AI model based on the context length and logical complexity of the call request.

[0034] The server is equipped with an AI model gateway. Before submitting an instruction, it can automatically route the call request to the most suitable AI model based on the context length and logical complexity of the call request. This AI model is the primary AI model, while other AI models are automatically used as backup AI models.

[0035] For example, the automatic routing rule is to route high-complexity "code / regular expression generation" tasks to strong inference models and simple "text case conversion" tasks to low-latency lightweight models, thereby achieving the optimal solution between response speed and computing cost.

[0036] S33: Select the AI ​​model to execute instructions and output structured results.

[0037] S34. Perform pattern consistency verification on the structured results, that is, verify whether the field types of the structured results conform to the input specifications of the front-end component (e.g., whether the pattern field is a string). This includes the following steps: S341. Perform noise filtering and format completion on the structured results; after obtaining the content returned by the primary AI model, automatically detect and remove common unstructured noise (such as Markdown code block markers json...). If the returned string contains extra characters at both ends, use regular expressions to precisely extract them.

[0038] S342. Extract the key structures from the structured results; determine whether the key structures conform to the tool's specific mode constraint specifications. If so, output the structured results and proceed to step S4 to send the verified results to the client's web page worker thread for isolated execution. Otherwise, proceed to step S343 to trigger the repair loop.

[0039] S343. The repair loop includes: when the critical structure does not conform to the tool-specific pattern constraint specification, an error log is output, and the error log, the original user instruction, and the tool-specific pattern constraint specification are input to the backup AI model; the backup AI model executes the instruction and outputs a structured result; and the structured result output by the backup AI model continues to be checked for pattern consistency.

[0040] Specifically, when the key structure does not conform to the tool-specific mode constraint specifications, a downgrade switch is triggered; when calling the backup AI model, the error log of the previous failure, the user's original instructions, and the tool-specific mode constraint specifications will be carried along and re-entered into the backup AI model to avoid the backup model repeating the same format error.

[0041] If all backup AI models fail to recover from the disaster, the "safe rollback" instruction is executed, which outputs a structured error object to the client, triggers the front-end user interface to release its loading state and restore it to the initial interactive interface.

[0042] In other embodiments of the present invention, if the primary AI model experiences an execution anomaly, fails to meet preset performance indicators, or triggers content security policies during the execution of instructions by the primary AI model (including but not limited to API response timeout (such as exceeding the preset 5000ms threshold) or triggering service rate limiting), a circuit breaker mechanism is automatically triggered to seamlessly and smoothly switch to the backup AI model.

[0043] Specifically, the circuit breaker mechanism generates an error code and then switches to a backup AI model. The backup AI model continues to execute the concatenated instructions based on the error code. If all backup AI models fail to recover from the disaster, a "safe rollback" instruction is executed, which outputs a structured error object to the client. Otherwise, the backup AI model outputs a structured result.

[0044] In other words, there are two timings for AI model switching: a repair loop and a circuit breaker mechanism. Specifically, when the structured result output by the primary AI model passes the pattern consistency check, the repair loop is triggered, and the model is switched. When the primary AI model fails to output a structured result due to execution anomalies, failure to meet preset performance indicators, or triggering content security policies, the circuit breaker mechanism is triggered, and the model is switched.

[0045] In some embodiments of the present invention, when switching models, the instruction template (System Prompt) can be automatically fine-tuned according to the characteristics of the alternative models to ensure that the task continues to flow without the user's awareness.

[0046] S4. The structured results are sent to the client's web page worker thread for isolated execution.

[0047] S41. Perform basic validity pre-checks (such as bracket matching checks) on the generated regular expressions or code snippets to prevent basic errors during the execution phase.

[0048] S42. Mount a global exception listener and timeout circuit breaker in the client sandbox to intercept exponential operations or infinite loops caused by uncontrollable divergent logic (commonly known as AI illusion), thereby preventing the main page rendering process from lag or crash.

[0049] For example, when finally executing the generated code snippet or regular expression matching, the execution flow is forcibly pushed to a separate worker thread in the user's terminal browser. By mounting a global exception listener and timeout circuit breaker (such as setting an execution limit of 200ms) in the front-end sandbox, it is ensured that any exponential operations or infinite loops caused by uncontrollable divergent logic are isolated and intercepted on the client side, and will absolutely not affect or cause lag or crashes in the Nuxt main page rendering process.

[0050] The technical solution of the present invention has the following technical effects compared with the prior art: This invention employs a "local priority" strategy to keep sensitive tasks processed on the user device, achieving "data not leaving the domain" and enhancing privacy and security. It utilizes multithreading and binary instruction sets for acceleration, addressing the impact of complex tasks on user interface responsiveness and optimizing performance. Through cloud-based artificial intelligence collaboration, it enables traditional static tools to understand natural language and process structured semantics, achieving intelligent boundary expansion.

[0051] Example 2, Reference Figure 2 and Figure 3 As shown, a cross-platform modular tool execution system 100 and an electronic device 200 are provided.

[0052] For the cross-platform modular tool execution system 100, refer to Figure 2 As shown, it includes: The request reading module 110 is used to read the user's tool call requests; this request reading module 110 is implemented through the user interface layer and is responsible for multimodal data input acquisition. This embodiment uses Nuxt4 as an example for illustration, but those skilled in the art should understand that the system of the present invention is also applicable to other modern front-end frameworks that support server-side rendering and edge computing (such as Next.js).

[0053] The path selection module 120 is used to determine the execution path based on the user selection or the tool's preset attribute labels; the execution path includes local edge execution path and cloud collaboration path.

[0054] The request processing module 130 is used to select an AI model based on the user request and output structured results. For the local edge execution path in the standard mode: high-performance, zero-data-upload local computation is implemented on the client side using WebAssembly and / or Web Worker (e.g., the Tesseract text recognition engine).

[0055] For the cloud-based collaborative path in the artificial intelligence model: Deployed on the server-side component, it is responsible for converting natural language into a structured DSL (Domain-Specific Language) and performing format auditing and schema (schema / definition) verification on the AI ​​output. This embodiment uses Nitro as an example for illustration, but those skilled in the art should understand that the system of this invention is also applicable to other modern front-end frameworks that support server-side rendering and edge computing (such as Next.js).

[0056] An isolated execution module 140 is used to execute the structured results on the client side. Similarly, this module is implemented through a user interface layer and is capable of dynamically rendering the results.

[0057] It should be understood that the cross-end modular tool execution system 100 here is embodied in the form of functional modules. The term "module" here can refer to application-specific integrated circuits (ASICs), electronic circuits, processors (e.g., shared processors, proprietary processors, or group processors, etc.) and memory for executing one or more software or firmware programs, integrated logic circuits, and / or other suitable components supporting the described functions. In an alternative example, those skilled in the art will understand that the cross-end modular tool execution system 100 can be specifically the electronic device 200 in the above embodiments, or the functions of the electronic device 200 in the above embodiments can be integrated into the cross-end modular tool execution system 100. The cross-end modular tool execution system 100 can be used to execute the various processes and / or steps corresponding to the electronic device 200 in the above method embodiments; to avoid repetition, these will not be described again here.

[0058] The aforementioned cross-platform modular tool execution system 100 has the function of implementing the corresponding steps executed by the electronic device 200 of the cross-platform modular tool execution method in Embodiment 1; the aforementioned functions can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the aforementioned functions. For example, the aforementioned acquisition module can be a communication interface, such as a transceiver interface.

[0059] In the embodiments of this application, Figure 2 The cross-end modular tool execution system 100 can also be a chip or a chip system, such as a system on chip (SoC).

[0060] Reference Figure 3 As shown, in this embodiment, an electronic device 200 is provided, including: Processor 210, and memory 220 and transceiver 230 communicatively connected to said processor; The memory 220 stores computer-executed instructions; the transceiver 230 is used for sending and receiving data. The processor 210 executes the computer execution instructions stored in the memory 220 to implement the calculation method in Embodiment 1.

[0061] It should be understood that the electronic device 200 can be used to perform the corresponding steps and / or processes in the above method embodiments. Optionally, the memory 220 may include read-only memory and random access memory, and provide instructions and data to the processor. A portion of the memory 220 may also include non-volatile random access memory. For example, the memory 220 may also store device type information. The processor 210 can be used to execute instructions stored in the memory 220, and when the processor 210 executes the instructions, the processor 210 can perform the corresponding steps and / or processes in the above method embodiments.

[0062] It should be understood that, in the embodiments of this application, the processor 210 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0063] In implementation, each step of the above method can be completed by the integrated logic circuitry of the hardware in the processor 210 or by instructions in software form. The steps of the method disclosed in the embodiments of this application can be directly embodied in the execution by the hardware processor, or by a combination of hardware and software modules in the processor 210. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory, and the processor executes the instructions in the memory, combining them with its hardware to complete the steps of the above method. To avoid repetition, detailed descriptions are not provided here.

[0064] Example 3 In this example, a computer-readable storage medium is provided. The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the cross-end modular tool execution method in Embodiment 1.

[0065] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0066] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0067] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0068] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0069] In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0070] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A cross-platform modular tool execution method, characterized in that, Includes the following steps: Receive tool invocation requests initiated by users on the client; Obtain the preset attribute tags of the invoked tool, and route the invocation request to the local edge execution path or the cloud collaboration path according to the preset attribute tags; The local edge execution path includes: initializing the web page assembly technology module and / or web page worker thread on the client; performing local computation using the client's local computing resources and outputting the results; The cloud-based collaborative path includes: the server executes instruction concatenation, sends the concatenated instructions to the AI ​​model to generate a structured result, performs pattern consistency verification on the structured result, and sends the verified result to the client's web page worker thread for isolated execution.

2. The cross-end modular tool execution method of claim 1, wherein, The server-side instruction concatenation includes: structurally concatenating a preset instruction template, tool-specific mode constraint specifications, user input context, and original natural language description to forcibly constrain the output boundaries of the AI ​​model.

3. The cross-end modular tool execution method of claim 1, wherein, The AI ​​model generates structured results by: determining the primary AI model based on the context length and logical complexity of the call request; executing instructions using the primary AI model; capturing the abnormal state if the primary AI model fails to execute, fails to meet preset performance indicators, or triggers content security policies; and then switching to a backup AI model, which continues to execute the concatenated instructions based on the abnormal state. If all backup AI models fail to recover from the disaster, a "safe rollback" instruction is executed, i.e., a structured error object is output to the client; otherwise, the backup AI models output structured results.

4. The cross-end modular tool execution method of claim 1, wherein, The pattern consistency verification includes: after filtering noise and completing the format of the structured result; extracting the key structure in the structured result; determining whether the key structure conforms to the tool's specific pattern constraint specification; if so, outputting the structured result; otherwise, triggering a repair loop.

5. The cross-end modular tool execution method of claim 4, wherein, The repair loop includes: when the critical structure does not conform to the tool-specific pattern constraint specification, outputting an error log; inputting the error log, the user's original instruction, and the tool-specific pattern constraint specification into the backup AI model; the backup AI model executing the instruction and outputting a structured result; and continuing to perform pattern consistency verification on the structured result output by the backup AI model.

6. The cross-end modular tool execution method of claim 1, wherein, The isolation execution of the web page worker threads sent to the client includes: mounting a global exception listener and a timeout circuit breaker in the client sandbox to intercept exponential operations or infinite loops caused by uncontrollable divergent logic, thereby preventing the main page rendering process from lag or crashing.

7. The cross-end modular tool execution method of claim 1, wherein, After receiving a tool call request initiated by the user on the client, the system also reads the user's mode selection, which includes a standard mode and an artificial intelligence mode; the standard mode is the local edge execution path, and the artificial intelligence mode is the cloud collaboration path.

8. A cross-end modular tool execution system for implementing the cross-end modular tool execution method according to any one of claims 1-7, characterized in that, include: The request reading module is used to read the user's tool call requests; The path selection module is used to determine the execution path based on user selection or the tool's preset attribute labels; The request processing module is used to select an AI model based on the user's request and output structured results. An isolated execution module is used to execute the structured results on the client side.

9. An electronic device, comprising: include: A processor, and a memory and a transceiver communicatively connected to the processor; The memory stores computer-executed instructions; the transceiver is used for sending and receiving data. The processor executes computer execution instructions stored in the memory to implement the cross-end modular tool execution method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the cross-end modular tool execution method as described in any one of claims 1-7.

Images