Interface: SecurityEvaluationResult

Defined in: .source/wunderland/src/security/types.ts:212

Overall security evaluation result.

Properties

action

action: GuardrailAction

Defined in: .source/wunderland/src/security/types.ts:214

Guardrail action to take

---

auditResult?

optional auditResult?: AuditResult

Defined in: .source/wunderland/src/security/types.ts:220

Dual-LLM audit result

---

classification?

optional classification?: InputClassificationResult

Defined in: .source/wunderland/src/security/types.ts:217

Pre-LLM classification result

---

metadata?

optional metadata?: Record<string, unknown>

Defined in: .source/wunderland/src/security/types.ts:235

Metadata for logging/observability

---

modifiedText?

optional modifiedText?: string

Defined in: .source/wunderland/src/security/types.ts:232

Modified text (if sanitization applied)

---

reason?

optional reason?: string

Defined in: .source/wunderland/src/security/types.ts:226

Combined reason for action

---

reasonCode?

optional reasonCode?: string

Defined in: .source/wunderland/src/security/types.ts:229

Reason code for programmatic handling

---

signedOutput?

optional signedOutput?: SignedAgentOutput

Defined in: .source/wunderland/src/security/types.ts:223

Signed output (if signing enabled)