Context and Token Budgets for School Moral Education and Class Teacher Work: Compare Documents and Locate Differences

Pastoral Education: review context & tokens for document checks. Check source versions, input mappings and the values received by downstream nodes.

What this category's data looks like

This page addresses routine material verification for school moral education and homeroom teacher work. Common materials include three types: themed class meeting plans, student comprehensive quality records, and home visit records. Themed class meeting plans are usually updated weekly or by theme, and contain activity objectives, procedures, hosts, and signature fields; student comprehensive quality records are updated by semester or stage, and contain items and quantity statistics such as ideological and moral character, academic performance, and activity participation; home visit records are updated per visit, and contain time, subject, key communication points, parent receipt, and teacher signature. When the three types of materials enter this function, themed class meeting plans and home visit records often serve as the text to be verified, while comprehensive quality records and ledgers serve as the comparison source, used to locate discrepancies in names, quantities, and dates for the same object across the two materials.

Configuration and implementation checks

This part is executed by application settings (response limit, etc.), which output the length of a single answer and the range of context that can be accommodated. The discrepancy list is then completed by the model, and the entry point and authorization are handled by the publishing service and file permissions.

ObjectCheckVerification action
Themed class meeting planThe class meeting name, meeting date, number of participants, and signature fields in the plan must be aligned item by item with the ledgerIn application settings, check the position of the response limit slider, and confirm that the values of maxResponse and maxContext can accommodate the entire plan and ledger comparison
Student comprehensive quality recordThe student names, number of items, record dates, and teacher signatures in the record must be consistent with the ledgerUse one real record to compare the two materials, check whether quoteMaxToken is sufficient to hold the quoted text, and record the number of array items separately
Home visit recordThe home visit time, subject name, receipt signature, and ledger fields must be consistentAsk a question that requires a long answer, see whether the ending is complete, and then write back the discrepancy fields
Discrepancy listWhether the names, quantities, and dates of the same object in the two materials matchList the inconsistent fields one by one, submit them to the reviewer for confirmation, and write them back

Three common errors

  • Symptom: The same themed class meeting plan has mismatched names or dates in two places. Cause: Versions are not synchronized, old drafts and new drafts are mixed, and there is no unified write-back.
  • Symptom: The number in the student comprehensive quality record is inconsistent with the ledger. Cause: After items are added or deleted, updates are not synchronized, and the quantity field remains at the old version.
  • Symptom: The home visit record lacks a signature or receipt. Cause: The material circulation step did not collect everything, and verification only looked at the body text without checking the signature field.

How to verify the setup

This function's acceptance check: ask a question that requires a long answer, see whether the ending is complete, and confirm that the response limit will not truncate in the middle of a sentence.

Business test: use one real record to find fields inconsistent with the ledger, record the name, quantity, or date discrepancies, and write them back.

  • In application settings, check the response limit switch and slider position, and confirm that the single-answer length matches the page requirements.
  • Compare usage from two runs, observe the change in single-run usage after adjusting the limit, and then decide whether to increase or decrease it.