Agent Armor API Reference
    Preparing search index...

    Class PatternDetector

    Generic detector driven by the pattern database. Replaces all hardcoded detector classes for pattern-based detection.

    Hierarchy (View Summary)

    Index
    • Parameters

      • opts: {
            category: TrapCategory;
            foldInside?: boolean;
            id: string;
            maskEvidence?: boolean;
            name: string;
            patterns: PatternEntry[];
            redactCredentials?: boolean;
            replaceText?: string;
            sanitizeMode?: "none" | "remove" | "replace";
            trapType: TrapType;
        }
        • category: TrapCategory
        • OptionalfoldInside?: boolean

          Scan the text as given and fold Unicode here instead: the tool-call reader must see the original characters. Patterns on the whole text run on a folded copy and map back; patterns on tool-call commands run on folded commands.

        • id: string
        • OptionalmaskEvidence?: boolean

          Redact matched text in Threat.evidence. Set for detectors whose matches are secrets, so a scan can't leak what it just found.

        • name: string
        • patterns: PatternEntry[]
        • OptionalredactCredentials?: boolean

          Cut credentials inside the matched text down to a prefix, keeping the rest.

        • OptionalreplaceText?: string
        • OptionalsanitizeMode?: "none" | "remove" | "replace"
        • trapType: TrapType

      Returns PatternDetector

    category: TrapCategory

    Which trap category this detector addresses

    id: string

    Unique identifier for this detector

    name: string

    Human-readable name

    trapType: TrapType
    • Last chance to alter matched text before it becomes Threat.evidence.

      Evidence travels: into ScanResult, the CLI's JSON/SARIF output, audit records, and whatever the caller logs. For most detectors the matched text IS the finding and must be preserved verbatim. For a detector whose matches are themselves secrets, echoing them would leak the credential into exactly the CI logs and audit trails the scan was meant to protect. Override to redact; the default is identity.

      Sanitization is unaffected — it works from Threat.location, not evidence.

      Parameters

      • match: string

      Returns string

    • Return sanitized content with threats neutralized. content is the original scanned text, never text another detector has already edited, so the threats' offsets are valid in it. Each detector's change is merged with the others' and applied once.

      Parameters

      • content: string
      • threats: Threat[]

      Returns string