What this category of data looks like
Data sources for this category include defect management systems, test reports, code commit records, and user feedback. Update frequency is daily or per version release cycle. Daily produced documents include defect tickets, regression test checklists, repair verification records, and defect statistical reports. Defect tickets contain fields such as defect ID, priority, reproduction steps, and associated versions. Regression checklists link to corresponding test cases and pending verification defects. Repair verification records mark verification results and executors. Data circulates through the full lifecycle of functional defects, and is compatible with FastGPT v4.17.0.
What constraints do these characteristics impose on the citation sources and traceability link
This category’s data is multi-source and circulates around defect IDs. Citation traceability requires binding to specific defect IDs, test case IDs, or version identifiers to avoid confusing defect data from different cycles. Document formats from different sources vary significantly. Configure unified citation marking rules separately for defect tickets, test reports, and code commits. Data updated daily or per version requires carrying version numbers in traceability information to ensure cited content matches the current test or repair version. Additionally, the time-consuming nature of defect reproduction and regression verification requires traceability information to directly link to specific execution results of verification records, to avoid vague citations leading to missed high-priority defects or version delays.
How to set the configuration
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
quoteMaxToken | 800–1200 token | Matches the average content length of a single defect ticket or regression checklist, to avoid truncating critical traceability information |
responseData | Only return fields associated with defect IDs and version numbers | Aligns with the core defect-ID-centric flow characteristics of this category of data, filtering out irrelevant content |
detail | Enable full traceability fields | Ensure citations include information required for compliance review, such as reproduction steps and verification results |
| Citation marking rules | Uniformly use the "[Defect ID:XXX]" format | Matches the field naming conventions of defect management systems, facilitating subsequent associated queries |
| Single traceability recall limit | Top 3 entries | Aligns with the conventional number of defects linked to a single regression test, avoiding information overload |
| Maximum citation mark length | 120 characters | Prevent overly long marks in output from negatively impacting reading experience |
Three Common Misconfigurations
- Phenomenon: Generated citation marks do not link to real defect IDs, and forged random IDs appear. Cause: Failed to bind to actual ID fields of the defect management system, and used generic placeholders directly to generate citations.
- Phenomenon: Unparsed text "Citation Mark: [1]" appears in the output content. Cause: Failed to turn off the automatic generation switch for generic citation marks, and did not configure exclusive defect ID format replacement rules for this category.
- Phenomenon: Too many irrelevant code commit records are returned in traceability results. Cause: Did not limit the return scope of
responseData, and did not filter code commit data unrelated to defects.
How to Confirm Configuration is Correct
- Import a real defect ticket and regression checklist, trigger the review process, and check whether the generated citations link to real defect IDs and version numbers.
- Review the output content, confirm that the generic "Citation Mark: [X]" format does not appear, and only the custom defect ID format is displayed.
- Check the number of traceability results, confirm that it does not exceed the set single recall limit.
- Check the configuration status of the
detailparameter, confirm that full field return is enabled, including reproduction steps and verification results.
The values given are common starting points and should be measured against the reader's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-10-19.