What this category's data looks like
Materials from peer review organizations usually come from review invitations, expert review comments, and author revision notes. Review invitation records record the invitation date, expert identity, and receipt status; expert review comments record the comment body, submission date, and signature; author revision notes record revision items, corresponding dates, and author signatures. Materials are updated by manuscript cycle, and the same manuscript may go back and forth for multiple rounds, so fields are prone to drift in names, quantities, dates, and signatures. When entering exception tracing and manual review, review invitations are used to confirm whether experts were effectively invited, expert review comments are used to compare comment items and signatures, and author revision notes are used to check revision items and receipts. The three types of materials jointly support the exception list and review conclusions, and the model answers based on them; external access and authorization are handled by the publishing service and file permissions.
Configuration and implementation checks
This part is executed by the memory turns and prompt configuration in the app settings, and the output references the historical message range and answer constraints. The exception list and review conclusions are then completed by the model. Memory turns control how many previous dialogue turns can be referenced at most when answering; if the value is too small, the preceding text cannot be used, and if the value is too large, irrelevant content is brought in and consumption increases.
| Object | Check | Verification action |
|---|---|---|
| Review invitation | The invitation date, expert name, and receipt status must correspond item by item with the submission date and expert name in the expert review comments; if a signature or receipt is missing, mark it | Check the memory turns value in the app settings, run a trial with two consecutive questions, and check whether the second round can carry forward the reference from the first round; use messages and chatId to locate the historical range |
| Expert review comments | The number of comment items, submission date, and signature must be consistent with the number of revision items, dates, and signatures in the author revision notes; if the version is not synchronized, mark it | Retrieve the original records and compare names, quantities, and dates; use variables to record the comparison results and confirm that the exception is supported by original records |
| Author revision notes | Revision items and receipts must be traceable to the source materials of the review invitation and expert review comments; if a signature or receipt is missing, mark it | Trace the three materials back by manuscript cycle and check signatures and dates; compare the usage records of two runs to confirm that consumption changes with the turn adjustment |
Three common errors
- Symptom: the expert name and invitation date for the same record do not match between the review invitation and the expert review comments. Cause: the two entries were not synchronized, and the name or date fields are maintained separately.
- Symptom: the number of revision items in the author revision notes is inconsistent with the number of items in the expert review comments. Cause: the version was not synchronized, and the revision notes were generated based on an old version of the comments.
- Symptom: the review conclusion lacks the reviewer's signature or receipt. Cause: the signature and receipt fields were not completed as the materials circulated, and the exception tracing chain is broken.
How to verify the setup
This feature acceptance: ask two consecutive follow-up questions and see whether the second question carries forward the reference from the first; check the memory turns value in the app settings, and use messages and chatId to locate the referenced historical message range.
Business test: create an exception record, see whether it can be traced back to the source materials, and confirm that the exception is supported by original records and the reviewer's signature.
- Check the current memory turns value in the app settings, run a trial with two consecutive questions, and confirm that the second round can correctly carry forward the reference from the first round.
- Compare the usage records of two runs to confirm that consumption changes with the turn adjustment; use
variablesto record the comparison results.