What the data for this category looks like
Typical data for this category comes from central enterprise internal control departments, audit departments, finance departments, subordinate enterprises, and external audit institutions. Daily documents include internal control self-assessment reports, audit notices, audit working papers, issue lists, rectification ledgers, and rectification closure certificates. Data sources cover enterprise internal control systems, ERP, financial sharing systems, audit management systems, and procurement contract project systems. The update frequency is annual. Documents include structured fields and unstructured text. Some documents require association with cross-system supporting materials such as vouchers and contracts. Fields include document number, generation date, responsible entity, rectification status, and more.
What constraints do these characteristics impose on the citation sources and traceability link
Annual updated data sources mean the knowledge base must regularly synchronize full audit and rectification data. A single review only needs to call the latest data associated with the current batch. Multiple document types and cross-system data sources require the traceability function to adapt to different document structures, and accurately match system identifiers and document numbers. The coexistence of structured fields and unstructured text requires traceability to support both field-level precise matching and text fragment association. The need to associate cross-system supporting materials means the traceability link must establish index associations across multiple data sources, to prevent citations from being disconnected from actual business.
How to configure the settings
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
quoteMaxToken | 800–1200 token | Adapts to the context length of long documents such as internal control self-assessment reports and rectification ledgers, ensuring citation fragments are complete and not redundant |
responseData | Return only associated document metadata and fragments | Audit file review requires clear citation of the source system, document number and field, to avoid returning irrelevant redundant content |
detail | Enabled | Specific sources of each citation must be marked in the output, to meet the traceability supervision requirements of internal control audits |
| Maximum number of documents recalled per single request | Top 3–5 | Audit sample extraction requires precise matching; too many recalled documents will interfere with reviewer judgment |
| Knowledge base update cycle | Synchronize annually | Matches the annual update frequency of data in this category, and complies with the data source synchronization logic of FastGPT v4.17.0 |
Three common misconfigurations
- The generated rectification review documents include irrelevant historical internal control self-assessment report content. The cause is that the value range of
quoteMaxTokenis not limited, and document data beyond the current review batch is recalled. - Citation traceability cannot be associated with specific audit management system data. The cause is that
responseDatais not configured to return metadata mode, and only plain text fragments are returned without carrying system and document numbers. - Supporting materials for audit issue rectification cannot be accurately located. The cause is that the
detailconfiguration item is not enabled, and specific fields and document ID of the citation source are not marked in the output.
How to verify the configuration is correct
- Upload a standard audit working paper and rectification ledger, trigger the FastGPT v4.17.0 review process, and check whether the output carries the document's system identifier and number.
- View the workflow configuration panel, confirm that the value of
quoteMaxTokenis within the 800–1200 token range, to avoid recalling redundant documents. - Verify that each citation in the generated review report corresponds to a clear source document and field information.
- Simulate a review of rectification closure certificates, check that only system data associated with the current batch is recalled.
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 2026-09-28.