Tool management method and electronic device

Abstract

The present disclosure provides a tool management method, comprising: obtaining description information of a plurality of tools, wherein the tools are used for being called by a target model, and the target model is capable of inferring a tool to be called from the plurality of tools according to a task and the description information and outputting a calling instruction; determining a target function according to the description information, the target function being an executable function overlapping among the plurality of tools; and rewriting the description information of at least one of the plurality of tools according to the target function, so that the description information of the at least one of the plurality of tools after the rewriting indicates that the tool is not suitable for executing the target function. The present disclosure also provides an electronic device.

Description

Technical Field

[0001] This disclosure relates to the field of large models, and more particularly to a tool management method and electronic device. Background Technology

[0002] In related technologies, when a large model calls a tool to process a task, it needs to determine the target tool from multiple tools with similar functions. Because the descriptions of tools are similar, the large model is confused about tools with similar functions, making it difficult to determine the target tool from multiple tools.

[0003] Therefore, a new tool management method and electronic device are needed. Summary of the Invention

[0004] In view of this, the present disclosure provides a tool management method and electronic device in an attempt to solve or at least alleviate at least one of the above-mentioned problems.

[0005] According to a first aspect of this disclosure, a tool management method is provided, comprising: acquiring description information of a plurality of tools, wherein the tools are used for invocation by a target model, the target model being able to infer the tools to be invoked from the plurality of tools based on a task and the description information and output invocation instructions; determining a target function based on the description information, the target function being an executable function that overlaps among the plurality of tools; and rewriting the description information of at least one of the plurality of tools based on the target function, wherein after rewriting, the description information of at least one of the plurality of tools indicates that the tool is not suitable for executing the target function.

[0006] Optionally, in the method according to this disclosure, determining the target function based on the description information includes: clustering multiple tools to obtain a tool cluster based on the description information of the tools; and determining the target function based on the description information of the tools in the tool cluster.

[0007] Optionally, in the method according to this disclosure, the step of clustering multiple tools to obtain a tool cluster based on the tool's description information includes: clustering multiple candidate tools to obtain the tool cluster based on the tool name and capability description information in the tool's description information.

[0008] Optionally, in the method according to this disclosure, rewriting the description information of at least one of the plurality of tools according to the target function includes: determining a target tool that meets the matching conditions with the target function based on the plurality of tools; rewriting the description information corresponding to the tool other than the target tool among the plurality of tools to first description information, wherein the first description information indicates that the tool is not suitable for performing the target function.

[0009] Optionally, in the method according to this disclosure, determining the target tool that meets the matching conditions with the target function based on the plurality of tools includes: generating test cases based on the target function; determining matching scores corresponding to the plurality of tools respectively based on the test cases, the matching scores representing the degree of matching between the tool and the target function; and determining the target tool based on the matching scores.

[0010] Optionally, in the method according to this disclosure, the target tool includes a first tool and a second tool, and the step of determining the target tool that satisfies the matching condition with the target function based on the plurality of tools includes: determining a first type of input parameters and a second type of input parameters based on the input parameters of the plurality of tools corresponding to the target function respectively; determining the first tool based on the first type of input parameters and the target function; and determining the second tool based on the second type of input parameters and the target function.

[0011] Optionally, in the method according to this disclosure, the test cases include a first test case corresponding to the first type of input parameters and a second test case corresponding to the second type of input parameters. Determining the first tool based on the first type of input parameters and the target function includes: determining a first candidate tool from the plurality of tools, wherein the input parameters of the first candidate tool corresponding to the target function are the first type of input parameters; determining a first matching score corresponding to the first candidate tool based on the first test cases; and determining the first tool based on the first matching score.

[0012] Optionally, in the method according to this disclosure, determining the first type of input parameters based on the input parameters of the plurality of tools corresponding to the target function includes: determining the first type of input parameters based on the input parameter type, required parameters, and parameter semantic description corresponding to each tool.

[0013] Optionally, in the method according to this disclosure, determining the target tool that meets the matching conditions with the target function based on the plurality of tools includes: searching for a corresponding external tool based on the target function;

[0014] The target tool is determined based on the external tool and the plurality of tools.

[0015] According to a second aspect of this disclosure, an electronic device is provided, comprising: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors perform the method described above. Attached Figure Description

[0016] The above and other objects, features and advantages of this disclosure will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0017] Figure 1 This is a schematic diagram illustrating an application scenario of a tool management method according to an embodiment of the present disclosure;

[0018] Figure 2 A flowchart illustrating a tool management method according to an embodiment of this application is shown schematically;

[0019] Figure 3 A schematic diagram of overlapping executable functions according to one embodiment of this application is shown;

[0020] Figure 4 A schematic diagram illustrating rewritten description information according to an embodiment of this application is shown;

[0021] Figure 5 A schematic diagram of a target determination tool according to an embodiment of this application is shown;

[0022] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing a tool management method according to an embodiment of this application. Detailed Implementation

[0023] The embodiments of this disclosure will now be described with reference to the accompanying drawings. Various details of the embodiments of this disclosure are included to aid understanding and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0024] In the technical solutions disclosed herein, the collection, storage, use, processing, transmission, provision, disclosure, and application of data (including but not limited to user personal information) comply with the provisions of relevant laws and regulations, necessary confidentiality measures have been taken, and they do not violate public order and good morals.

[0025] Figure 1 This is a schematic diagram illustrating an application scenario of a tool management method according to an embodiment of this disclosure. It should be noted that... Figure 1 The examples shown are merely examples of application scenarios that can be applied to the embodiments of this disclosure, in order to help those skilled in the art understand the technical content of this disclosure, but do not mean that the embodiments of this disclosure cannot be used in other devices, systems, environments or scenarios.

[0026] like Figure 1As shown, application scenario 100 according to an embodiment of this application may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links or fiber optic cables. For example, a user can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send information, etc.

[0027] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be electronic devices such as smartphones and personal computers. Applications can run on these electronic devices, and these applications can include, but are not limited to, web browser applications, search applications, instant messaging tools, artificial intelligence agents, etc. (these are just examples). Furthermore, this application embodiment does not limit the form of the application, and includes, but is not limited to, applications installed on the terminal devices, mini-programs, etc., and can also be in the form of web pages.

[0028] Server 105 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud computing services, cloud storage, and network services. Server 105 can be a server providing various services. For example, server 105 may have a cloud-based large model deployed on it to process user requests sent by the artificial intelligence agent in the first terminal device 101, the second terminal device 102, and the third terminal device 103, and to feed back the processing results (such as files, information, or data obtained or generated according to the user request) to the terminal devices.

[0029] It should be noted that the tool management method provided in this application embodiment can generally be executed by terminal devices 101-103.

[0030] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0031] Figure 2 A flowchart illustrating a tool management method according to an embodiment of this application is shown schematically. Figure 2 As shown, the tool management method according to the embodiments of this application may include steps 210-230.

[0032] First, step 210 is executed to obtain description information of multiple tools. These tools are used for the target model to call. The target model can infer the tool to be called from multiple tools based on the task and description information and output the calling instruction.

[0033] According to one implementation, the target model can be implemented as the currently invoked large model, which can be one of the following: a Large Language Model (LLM), a multimodal large model, a model capable of generating text, a model with more than 100 million adjustable parameters, a model that can output corresponding results based on input natural language instructions, a model that can convert input natural language instructions into feature sequences and infer results based on the feature sequences, a model that can convert input information into feature sequences and infer subsequent features of the feature sequences, or a model that can generate corresponding results based on prompts. The tool is a Model Context Protocol (MCP) tool invoked when the large model processes tasks, such as a text processing tool, an image processing tool, or an operation execution tool. This application does not limit the specific type of the target model and the tool. Descriptive information refers to information describing what functions and tasks the tool can perform and what capabilities it possesses. The invocation instruction is the content in the text information generated by the model inference that indicates the invocation of the tool.

[0034] Subsequently, step 220 is executed to determine the target function based on the description information. The target function is an executable function that overlaps among multiple tools. Figure 3 A schematic diagram illustrating overlapping executable functions according to one embodiment of this application is shown. Figure 3 As shown, the scope of functions described in the description information of the first tool includes the first function and the second function; the scope of functions described in the description information of the second tool includes the first function and the third function; and the scope of functions described in the description information of the third tool includes the first function and the fourth function. If the executable functions of the first, second, and third tools overlap, and this first function is defined as the target function.

[0035] Subsequently, step 230 is executed, in which the description information of at least one of the multiple tools is rewritten according to the target function. After rewriting, the description information of at least one of the multiple tools indicates that the tool is not suitable for performing the target function.

[0036] According to one implementation, after rewriting, the data representing tools suitable for performing the target function in the description information can be reduced. For example, the scope of the function described by the description information of the first tool includes the first function and the second function; the scope of the function described by the description information of the second tool includes the first function and the third function; and the scope of the function described by the description information of the third tool includes the first function and the fourth function. After rewriting, the scope of the function described by the description information of the first tool includes the first function and the second function; the scope of the function described by the description information of the second tool includes the first function and the third function; and the scope of the function described by the description information of the third tool only includes the fourth function. Previously, when determining the tool corresponding to the target function, the model needed to select from the first tool, the second tool, and the third tool, but now it only needs to select from the first tool and the second tool; thereby narrowing the scope of tool determination, reducing the information that needs to be compared, and improving the efficiency of tool determination.

[0037] Figure 4 A schematic diagram illustrating rewritten description information according to an embodiment of this application is shown. For example... Figure 4 As shown, after rewriting the description information, the scope of the function described by the description information of the first tool includes the first function and the second function; the scope of the function described by the description information of the second tool includes the third function but does not include the first function; the scope of the function described by the description information of the third tool includes the fourth function but does not include the first function; after setting, the description information representation of the second tool and the third tool is not suitable for performing the first function.

[0038] This application modifies the description information of multiple tools when they have overlapping executable functions, so that the description information determines that the multiple tools do not have overlapping executable functions. This makes it easier for the target model to select a tool from multiple tools when it needs to call a tool to achieve the target function. Instead, it only needs to select the tool that achieves the target function and call it, avoiding confusion between multiple tools and improving the efficiency of tool calling.

[0039] According to one implementation, determining the target function based on description information includes: clustering multiple tools to obtain a tool cluster based on the tool description information; and determining the target function based on the description information of the tools in the tool cluster.

[0040] According to another implementation, determining the target function based on the description information includes: obtaining statistical data on the tool category tags included in the description information, and determining the target function. For example, the function corresponding to the highest value in the statistical data is determined as the target function.

[0041] According to another implementation, determining the target function based on the description information includes: determining the target function from multiple preset functions based on the matching relationship between the description information and multiple preset functions. The multiple preset functions can be important functions in a pre-defined scenario. By matching the description information of the multiple functions with the preset functions, it can be determined which preset function corresponds to multiple tools, thereby identifying that preset function as the target function.

[0042] According to one implementation, when clustering multiple tools based on their description information, the description information of the multiple tools can be directly input into a model with semantic analysis capabilities for processing to determine at least one cluster center; each cluster center represents a target function. Subsequently, the multiple tools are clustered based on the at least one cluster center to obtain a tool cluster. A tool cluster may include at least one tool; a tool cluster indicates that all tools in the cluster have a target function, and that multiple tools in the cluster have overlapping executable functions. Next, the description information of the tools in the tool cluster is input into a model with semantic analysis capabilities for semantic analysis to determine the target function.

[0043] According to one implementation, multiple tools are clustered to obtain a tool cluster based on the tool's description information, including: clustering multiple candidate tools based on the tool name and capability description information in the tool's description information to obtain a tool cluster.

[0044] According to one implementation, candidate tools refer to tools currently available in the large model. The descriptive information of a tool may include multiple pieces of information or multiple types of information, such as the tool's name and capability description information. The tool name serves as an identifier for the tool; the capability description information describes the tasks the tool can perform, its capabilities, and the functions it implements, and can be defined by the developer or previous users. Both the tool name and capability description information reflect the tool's function; the words in the tool name and capability description information can directly describe the tool's function, and can also describe the tool's type; the tool type can be divided according to function, such as recognition type, word segmentation type, etc. After determining the tool type, candidate tools can be clustered based on the functions possessed by tools of this common type.

[0045] According to one implementation, determining a target tool that meets matching conditions with a target function based on multiple tools includes: searching for a corresponding external tool based on the target function; and determining the target tool based on the external tool and multiple tools.

[0046] According to one implementation, the target model can currently access or register at least one tool, referred to as an internal tool. Relative to these tools, tools not accessed or registered in the target model can be searched based on the target function and designated as external tools. When determining the tool performing the target function from multiple tools in a set, only the target tool is identified as performing the target function. Searching for external tools expands the scope of this set. This ensures that even when more large models are accessed or registered in the future, the determination of the tool performing the target function remains the same, only the target tool is identified.

[0047] According to one implementation, rewriting the description information of at least one of a plurality of tools based on a target function includes: determining a target tool that meets the matching conditions for the target function based on the plurality of tools; rewriting the description information corresponding to tools other than the target tool among the plurality of tools into first description information, the first description information indicating that the tool is not suitable for performing the target function. The first description information includes at least one of the following: the description information does not contain information corresponding to the target function, or the description information describes that the tool cannot achieve the target function.

[0048] According to one implementation, the number of target tools can be one; after the target tool is determined, when it is necessary to call the tool to implement the target function, only the target tool is called based on the description information; other tools will not be called because their first description information does not include the target function, indicating that the tool is not suitable for performing the target function; thus avoiding confusion of the target model among multiple tools and improving the efficiency of tool calling.

[0049] According to one implementation, the number of target tools can be multiple. Before rewriting, the target function corresponds to multiple target function tools, and the number of target tools is less than the number of target function tools. Therefore, by rewriting the description information corresponding to tools other than the target tools among the multiple tools as the first description information, the number of tools corresponding to the target function is reduced, and the confusion between tools is also reduced accordingly.

[0050] According to one implementation, determining a target tool that meets matching conditions with the target function based on the plurality of tools includes: generating test cases based on the target function; determining matching scores for each of the plurality of tools based on the test cases, where the matching scores characterize the degree of matching between the tool and the target function; and determining the target tool based on the matching scores. For example, a matching condition represents the highest matching score.

[0051] According to one implementation method, test cases can be written based on the target function to determine the tool's performance in fulfilling that function, such as accuracy, efficiency, crash susceptibility, and error probability. Based on the tool's performance in fulfilling the target function, a matching score can be determined for each tool, and the tool with the highest matching score is identified as the target tool.

[0052] According to one implementation, determining a target tool that satisfies a matching condition with the target function based on the plurality of tools includes: determining a target priority for each tool relative to the target function based on preset priority information corresponding to the plurality of functions in the tool's description information, and determining the target tool from the tools based on the target priority. For example, the matching condition indicates that the target priority is the highest.

[0053] According to one embodiment, the target tool includes a first tool and a second tool. Determining the target tool that meets the matching conditions with the target function based on the plurality of tools includes: determining a first type of input parameters and a second type of input parameters based on the input parameters of the plurality of tools corresponding to the target function; determining the first tool based on the first type of input parameters and the target function; and determining the second tool based on the second type of input parameters and the target function.

[0054] According to one implementation, multiple categories of input parameters, including a first category of input parameters and a second category of input parameters, are determined based on the input parameters of multiple tools corresponding to the target function.

[0055] According to one implementation, the tool suitable for performing the target function, determined based on the description information, may have different input parameters, such as a first type of input parameters (e.g., a, b), a second type of input parameters (e.g., a, b, c), and a third type of input parameters (e.g., a). For example, for the same target function: querying the weather, the first type of input parameters might be the input time information and latitude and longitude coordinates, the second type of input parameters might be the input date and city name, and the third type of input parameters might be only the input city name. This application does not limit the specific type or number of input parameters.

[0056] According to one implementation, the target tool may include different tools with different input parameters. For example, a first tool may have a first type of input parameters, a second tool may have a second type of input parameters, and a third tool may have a third type of input parameters. Therefore, a corresponding tool can be determined based on different types of input parameters and target functions, so that a target tool can be determined for each type of input parameter to meet the tool invocation needs of users when inputting different parameters.

[0057] According to one implementation, the test cases include a first test case corresponding to a first type of input parameters and a second test case corresponding to a second type of input parameters. Determining a first tool based on the first type of input parameters and the target function includes: determining a first candidate tool from a plurality of tools, wherein the input parameters of the first candidate tool corresponding to the target function are first type of input parameters; determining a first matching score corresponding to the first candidate tool based on the first test cases; and determining the first tool based on the first matching score.

[0058] According to one implementation, when determining the corresponding tool based on the type of input parameters and the target function, candidate tools can be selected based on the type of input parameters. Then, the target tool is determined from these candidate tools, ensuring that one target tool is assigned to each type of input parameter, thus meeting the user's tool invocation needs under different input parameter conditions. When determining one target tool for each type of input parameter, corresponding test cases are determined based on that type of input parameter. For example, a first test case is set for the first type of input parameter, and a second test case is set for the second type of input parameter to meet testing requirements. The target tool of the corresponding type is then tested according to the corresponding type of test cases to select the appropriate target tool from the candidate tools of that type.

[0059] Figure 5 A schematic diagram of a target determination tool according to an embodiment of this application is shown; as follows: Figure 5 As shown, the target model can call upon the first to fifth alternative tools; the corresponding tools can be determined based on the first type of input parameters, including the first and second alternative tools; and the corresponding tools can be determined based on the third type of input parameters, including the third, fourth, and fifth alternative tools. Subsequently, the first and second alternative tools are tested according to the first test cases, determining that the target tool includes the first tool, which is the tool corresponding to the first type of input parameters and the target function; the first test cases are set according to the first type of input parameters, for example, all having the same first input parameters (a, b). The third, fourth, and fifth alternative tools are tested according to the second test cases, determining that the target tool includes the second tool, which is the tool corresponding to the second type of input parameters and the target function; the second test cases are set according to the second type of input parameters, for example, all having the same second input parameters (a, b, c).

[0060] According to one implementation, a first type of input parameters is determined based on the input parameters of multiple tools corresponding to the target function, including: determining the first type of input parameters based on the input parameter type, required parameters, and parameter semantic description corresponding to each tool.

[0061] According to one implementation, the input parameter type includes input parameters as integers, decimals, etc.; required parameters include whether the parameter is required or not; the parameter semantic description is a description of the parameter, defining the input format, purpose and specific meaning of the parameter, etc.

[0062] According to one implementation, tools with the same required parameters can be defined as tools corresponding to the same type of input parameters; tools with the same required parameters and tools with the same required parameters can also be defined as tools corresponding to the same type of input parameters.

[0063] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing a tool management method according to an embodiment of this application.

[0064] like Figure 6 As shown, an electronic device 400 according to an embodiment of this application includes a processor 401, which can perform various appropriate actions and processes according to a program stored in a read-only memory 402 or a program loaded from a storage portion 408 into a random access memory 403. The processor 401 may include, for example, a general-purpose microprocessor, an instruction set processor and / or an associated chipset and / or a dedicated microprocessor. The processor 401 may also include onboard memory for caching purposes. The processor 401 may include a single processing unit or multiple processing units for executing different steps of the method flow according to an embodiment of this application.

[0065] Random access memory 403 stores various programs and data required for the operation of electronic device 400. Processor 401, read-only memory 402, and random access memory 403 are interconnected via bus 404. Processor 401 executes various steps of the method flow according to embodiments of this application by executing programs stored in read-only memory 402 and / or random access memory 403. It should be noted that the programs may also be stored in one or more memories other than read-only memory 402 and random access memory 403. Processor 401 may also execute various steps of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0066] According to embodiments of this application, the electronic device 400 may further include an input / output interface 405, which is also connected to a bus 404. The electronic device 400 may also include one or more of the following components connected to the input / output interface 405: an input section 406 including a keyboard, mouse, etc.; an output section 407 including a cathode ray tube, liquid crystal display, etc., and a speaker, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card, such as a local area network card, modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output interface 405 as needed. A removable medium 411, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 410 as needed so that computer programs read from it can be installed into the storage section 408 as needed.

[0067] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0068] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A tool management method, comprising: Obtain description information of multiple tools, wherein the tools are used for invocation by the target model, and the target model can infer the tool to be invoked from the multiple tools based on the task and the description information and output the invocation command; The target function is determined based on the description information, and the target function is an executable function that overlaps among the plurality of tools; Based on the target function, the description information of at least one of the plurality of tools is rewritten. After rewriting, the description information of at least one of the plurality of tools indicates that the tool is not suitable for performing the target function.

2. The method as described in claim 1, wherein, Determining the target function based on the described information includes: Based on the description information of the tools, multiple tools are clustered to obtain tool clusters; The target function is determined based on the description information of the tools in the tool suite.

3. The method as described in claim 2, wherein clustering multiple tools to obtain a tool cluster based on the tool's description information includes: The tool cluster is obtained by clustering multiple candidate tools based on the tool name and capability description information in the tool description information.

4. The method of claim 1, wherein rewriting the description information of at least one of the plurality of tools according to the target function includes: Based on the multiple tools, a target tool that meets the matching conditions for the target function is determined; The description information corresponding to the tools other than the target tool among the plurality of tools is rewritten as the first description information, which indicates that the tool is not suitable for performing the target function.

5. The method according to claim 4, wherein determining the target tool that satisfies the matching condition with the target function based on the plurality of tools includes: Generate test cases based on the target functionality; Based on the test cases, a matching score is determined for each of the multiple tools, and the matching score represents the degree of matching between the tool and the target function. The target tool is determined based on the matching score.

6. The method according to claim 4 or 5, wherein the target tool comprises a first tool and a second tool, and the step of determining the target tool from the plurality of tools that satisfies the matching condition for the target function comprises: The first type of input parameters and the second type of input parameters are determined based on the input parameters of the multiple tools corresponding to the target function, respectively. The first tool is determined based on the first type of input parameters and the target function; The second tool is determined based on the second type of input parameters and the target function.

7. The method according to claim 6, wherein the test cases include a first test case corresponding to the first type of input parameters and a second test case corresponding to the second type of input parameters, and the step of determining the first tool based on the first type of input parameters and the target function includes: A first candidate tool is determined from the plurality of tools, wherein the input parameters of the first candidate tool corresponding to the target function are the first type of input parameters; Determine the first matching score corresponding to the first candidate tool based on the first test case; The first tool is determined based on the first matching score.

8. The method according to claim 6, wherein determining the first type of input parameters based on the input parameters of the plurality of tools corresponding to the target function respectively includes: The first type of input parameters is determined based on the input parameter type, required parameters, and parameter semantic description corresponding to each tool.

9. The method according to claim 4, wherein determining a target tool that satisfies the matching condition with the target function based on the plurality of tools, includes: Search for corresponding external tools based on the target function; The target tool is determined based on the external tool and the plurality of tools.

10. An electronic device, comprising: One or more processors; Storage device for storing one or more programs. Wherein, when the one or more programs are executed by the one or more processors, the one or more processors perform the method according to any one of claims 1 to 8.

Images