What this category's data looks like
This page focuses on kindergarten game activity organization. The material sources are class observation records, area activity plans, material placement lists, activity reflection forms, and next week's plan forms. Observation records are filled in per instance, area plans are updated weekly, material lists are adjusted along with placement, and reflections and next week's plans are connected on weekends. When entering the Q&A acceptance and missing-item follow-up function, the correspondence between observation records and area plans is used to determine whether activity goals have been achieved; the quantity of materials placed is used to check whether materials are sufficient; the conversion from reflection to next week's plan is used to track whether the plan has been updated. Common inconsistencies include insufficient materials, goals not matching observations, and plans not updated. The acceptance records and the completed follow-up questions are output by the tool calling node as the calling result, and then handed over to subsequent AI nodes or manual verification for completion.
Configuration and implementation checks
This part is executed by the tool calling node in the workflow (the model decides whether to call it, and the node executes it), outputting the tool calling result (the input parameters and return in the calling record). "Acceptance records and completed follow-up questions" are then completed by subsequent AI nodes or manual verification. The model decides whether to call the tool this time based on the tool description, parameter descriptions, and whether parameters are required; when there are no required parameters, the model may fabricate parameter values on its own, and when there are required parameters but no suitable value in the context, it may not call.
| Object | Check | Verification action |
|---|---|---|
| Correspondence between observation records and area plans | Whether the achievement of activity goals is consistent with the observed behavior | Open the description field of the tool node and check whether the description can be understood on its own; run a test with a question missing the goal field to see whether it first asks a follow-up question instead of directly fabricating |
| Quantity of materials placed | Whether a follow-up question to complete the missing item is triggered when materials are insufficient | Open each parameter configuration one by one and check the two items: description and whether it is required; run a test with a question missing the quantity field to confirm that the returned result does not contain values that do not exist in the business materials |
| Conversion from reflection to next week's plan | Whether it identifies and asks a follow-up question when the plan has not been updated | On the canvas, check the diamond-shaped connections and node status, and confirm that the tool calling node connection is connected to the target tool; check the settings related to toolChoice and functionCall |
| Tool calling node connections | Whether results cross over into each other after multiple tools are connected in series | On the canvas, check the diamond-shaped connections and node status, and confirm that toolChoice, functionCall, and the node status are consistent |
Three common errors
- Phenomenon: The model did not call the tool and directly gave an answer. Cause: The tool description does not state when it should be used, or the parameters are marked as required but the prompt does not guide follow-up questions.
- Phenomenon: The returned result contains values that do not exist in the business materials. Cause: When there are no required parameters, the model fabricates parameter values on its own, without corresponding to the material fields.
- Phenomenon: Insufficient materials, goals not matching observations, and plans not updated are not identified. Cause: The two items, parameter description and whether it is required, were not checked one by one, and missing items were not followed up on first.
How to verify the setup
Acceptance of this function: run a test with a question missing required parameters, check the input parameters and return in the calling record, and confirm whether it asks a follow-up question or calls according to the description; check whether toolChoice and functionCall are consistent with the node status.
Business test: ask a question missing key information and see whether it first asks a follow-up question before answering; if it does not ask a follow-up question, check whether the two items, parameter description and whether it is required, correspond to the business material fields.
- Open the description field in the tool node, check whether this description can be understood when taken out on its own, and confirm that the diamond-shaped interface is connected to the target tool.
- Run a test with a question missing that field to see whether it first asks a follow-up question instead of directly fabricating; check whether the returned result contains values that do not exist in the business materials.