SafetyFinding
Models & Types · AiDotNet.Safety
Represents a single safety finding from a safety module evaluation.
For Beginners
When the safety system checks your content, it may find zero or more problems. Each problem is represented as a SafetyFinding with details about what was detected and how serious it is.
How It Works
A safety finding captures one specific issue detected during safety analysis, including what was found, how severe it is, where it was found, and what action is recommended.
Constructors
| Constructor | Summary |
|---|---|
SafetyFinding() |
Properties
| Property | Summary |
|---|---|
SafetyAction RecommendedAction { get; set; } |
Gets the recommended action for this finding. |
SafetyCategory Category { get; set; } |
Gets the safety category of this finding. |
SafetySeverity Severity { get; set; } |
Gets the severity level of this finding. |
double Confidence { get; set; } |
Gets the confidence score for this finding (0.0 to 1.0). |
int SpanEnd { get; set; } |
Gets the end offset (exclusive) of the finding in the input. |
int SpanStart { get; set; } |
Gets the start offset (character index or sample index) of the finding in the input. |
string Description { get; set; } |
Gets a human-readable description of the finding. |
string Excerpt { get; set; } |
Gets the excerpt of content that triggered this finding. |
string SourceModule { get; set; } |
Gets the name of the safety module that produced this finding. |