Graphorin API reference v0.8.0
Graphorin API reference / @graphorin/tools / / InboundSanitizationWithClassifierOptions
Interface: InboundSanitizationWithClassifierOptions
Defined in: packages/tools/src/inbound/sanitize.ts:211
Options for applyInboundSanitizationWithClassifier.
Stable
Properties
| Property | Modifier | Type | Description | Defined in |
|---|---|---|---|---|
body | readonly | string | - | packages/tools/src/inbound/sanitize.ts:212 |
budgetMs? | readonly | number | - | packages/tools/src/inbound/sanitize.ts:219 |
classifier? | readonly | InjectionClassifier | B4 (D-12): optional pluggable injection classifier consulted AFTER the regex pass, on the already-sanitized body. A flagged verdict appends classifier:<id> to patternsHit (audit signal); the body is never modified by the classifier (it names no spans to strip). Errors are swallowed by the resilience contract - the regex outcome stands alone. | packages/tools/src/inbound/sanitize.ts:228 |
contentOrigin? | readonly | string | - | packages/tools/src/inbound/sanitize.ts:216 |
failClosed? | readonly | boolean | - | packages/tools/src/inbound/sanitize.ts:217 |
patterns? | readonly | readonly ImperativePattern[] | - | packages/tools/src/inbound/sanitize.ts:218 |
policy | readonly | InboundSanitizationPolicy | - | packages/tools/src/inbound/sanitize.ts:213 |
toolName | readonly | string | - | packages/tools/src/inbound/sanitize.ts:215 |
trustClass | readonly | ToolTrustClass | - | packages/tools/src/inbound/sanitize.ts:214 |