What this category's data looks like
This page addresses the kindergarten home-school co-education communication scenario, with materials drawn from daily conversation records, notifications, and feedback. The category verification dimensions include committed items and response times, descriptions of children's performance, classification of parents' requests, and notification delivery. When these materials enter the conversation record interface, each plays a different role: committed items and response times are used to verify whether the loop is closed, descriptions of children's performance are used to compare against observation records, classification of parents' requests is used to determine whether feedback is categorized, and notification delivery is used to confirm reach. The update cadence rolls by class and time period, and field characteristics include application and session identifiers, role fields, citation and latency fields. Conversation records are pulled in pages by application and session dimensions, returning two role types, questions and answers, with citations and latency. After materials enter this feature, session evidence is first formed, and then manually or by an external system it is registered as a handover-ready deployment status and operation record.
Configuration and implementation checks
This step is executed by the conversation record interface (by application and session dimensions), outputting a session message list and total record count, including role and citation information. "Handover-ready deployment status and operation records" are then completed manually or by an external system. This interface only retrieves session evidence and does not write to business ledgers.
| Object | Check | Verification action |
|---|---|---|
| Application and session identifiers | Consistent with the actual application and actual session; committed items and response times are traceable | Check the source of chatId and value in the request parameters, and confirm they correspond to the current application and current session |
| Pagination parameters | The number of records matches the total record count; classification of parents' requests is not omitted | Pull two pages consecutively and check the total record count and advance the offset by the number of records retrieved, confirming response code of 200 and no duplicates |
| Role fields | message roles correspond to the original conversation; descriptions of children's performance can be located | Look at the order of the two role types in the returned list, and sample one record to check the role and citation information |
| Citation and latency fields | A single record shows citations and latency; notification delivery can be checked | Open one record to check citations and latency; record the citation count separately by array length |
Three common errors
- Symptom: A committed item appears in the conversation but there is no subsequent response record. Cause: Only the first page was retrieved and subsequent pages were not pulled, causing the total record count to be inconsistent with the interface session.
- Symptom: Descriptions of children's performance do not match observation records. Cause: The message list was treated as a business material version ledger, and records were not compared one by one by role and citation information.
- Symptom: Classification of parents' requests is missing or feedback is not categorized. Cause: The interface returned empty but there is a session on the interface, and the source of the application and session identifiers was not checked.
How to verify the setup
This feature acceptance: page through by total record count total to retrieve everything, sample one record to check role, citation information, and time, confirm response code of 200, and check the message identifiers and content fields _id, dataId and value.
Business test: check each item once according to the handover form, see which record is missing, and confirm that version, connection, and task record are consistent.
- Pull two pages consecutively and check the total record count and advance the offset by the number of records retrieved, confirming no duplicates and response
codeof 200. - Open one record to check role, citations, and latency; record the citation count separately by array length, and check the citation limit against the tokens of the citation text.