What this category of data looks like
The core data for this category includes security training courseware, phishing drill plans, phishing email templates, and drill result reports. Common data sources are employee contact lists, email gateways, internal training platforms, and user click statistics systems. Data update frequency is mostly quarterly or annual. Document structures include drill audience rules, template parameters, and result statistics fields. There are no uniform mandatory units, and fields are adjusted on demand based on individual drill scenarios.
What constraints do these characteristics impose on the citation and traceability link
The data characteristics of this category impose multiple constraints on the citation and traceability process. Dispersed multi-source data requires traceability to link original records from multiple systems such as employee contact lists and email gateways, ensuring that drill audience and click data can be cross-verified. Low update frequency requires that referenced source data be locked to the version corresponding to the target cycle, preventing traceability failure from subsequent data updates. Diverse document types require matching different traceability fields: training courseware must be linked to training plan numbers, and email templates must be bound with version identifiers. Severe consequences of errors require that traceability information be fully retrievable, with no missing key evidence nodes.
How to configure settings
This document is written based on FastGPT v4.17.0.
| Configuration Item | Recommended Approach | Rationale |
|---|---|---|
| Maximum number of single retrieval entries | Top 3-5 entries | The number of core documents for this category is limited, and excessive entries will distract from traceability focus |
quoteMaxToken | 1200-1500 token | Matches the common document chunk length for this category, avoiding truncation of key traceability fields |
| Citation source address display | Enabled | Full display of source document paths is required to meet compliance requirements for audit traceability |
detail | full | Retain all traceability information from source documents, including drill numbers and version identifiers |
| Maximum context length | 2000-2500 token | Integrate traceability content from multiple sources to support cross-system cross-verification |
responseData | Include source document metadata | Ensure traceability information includes metadata such as document creation time and update cycle, matching the traceability requirements of low update frequency |
Three common configuration mistakes
- Phenomenon: The full address of the source file is not shown in the response. Cause: The citation source address display configuration item is not enabled, resulting in missing traceability information.
- Phenomenon: Eligible document chunks are retrieved, but the actual number of returned citations exceeds the set limit. Cause: The matching relationship between
quoteMaxTokenand the maximum number of single retrieval entries is not correctly configured, causing chunks that exceed the length limit to still be included in the results. - Phenomenon: Referenced traceability fields are incomplete, with missing drill numbers or version identifiers. Cause:
detailis not configured to full mode, only partial document content is returned, failing to meet audit requirements.
How to confirm configuration is complete
- Initiate an API Q&A targeting phishing drill plans or training courseware, check whether the full path of the source document is included in the response.
- Review the number of citations in the response, confirm it matches the configured maximum number of single retrieval entries.
- Verify the fields in the referenced content, confirm they include traceability information such as drill numbers and version identifiers.
- Adjust the
quoteMaxTokenparameter, verify that the length of referenced content in the response matches the set value.
The values provided 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.