Expunct

Expunct Sanitize Text

Redact PII from text and return the sanitized result immediately. Detects and replaces personally identifiable information (PII) in the provided text. Supports 27+ entity types across PII (names, SSNs, addresses), PCI (credit cards, bank accounts), and PHI (medical licenses) categories. This is synchronous — results return immediately (~100ms for typical text). Args: params (SanitizeTextInput): Validated input parameters: - text (str): Text to redact (1

Expunct Sanitize File

Redact PII from a file (document, image, video, or audio). Creates a background redaction job for the file. Use expunct_check_status to poll for completion, then expunct_get_findings to see what was detected. Supported formats: - Documents: PDF, DOCX, TXT, MD - Images: PNG, JPG, BMP, TIFF - Video: MP4, AVI, MKV, MOV, WEBM - Audio: WAV, MP3, FLAC, OGG, M4A Local file paths work in stdio mode. Cloud URIs (gs://, s3://, https://) work in any mode. Args:

Expunct Sanitize Batch

Redact PII from multiple files in a single batch operation. Creates a batch job containing individual redaction jobs for each file. Use expunct_check_status with any job_id to track individual progress. Args: params (SanitizeBatchInput): Validated input parameters: - paths (list[str]): File paths or cloud URIs (1-100 items) - pii_types (list[str] | None): Entity types to detect - language (str): 'en' or 'es' Returns: str: JSO

Expunct Check Status

Check the status and progress of a redaction job. Use this after calling expunct_sanitize_file or expunct_sanitize_batch to track job progress. Poll periodically until status is 'completed' or 'failed'. Args: params (CheckStatusInput): Input parameters: - job_id (str): Job ID from sanitize_file or sanitize_batch Returns: str: JSON with schema: { "job_id": str, "status": str, # "pending", "processing", "comp

Expunct List Entity Types

List all supported PII entity types grouped by category. Returns the complete list of entity types that can be detected, organized by category (PII, PCI, PHI). Use these type names with the pii_types parameter in sanitize_text, sanitize_file, and sanitize_batch. Returns: str: JSON with schema: { "categories": { "PII": ["PERSON", "PHONE_NUMBER", ...], "PCI": ["CREDIT_CARD", ...], "PHI": ["MEDICAL_LICENSE

Expunct Get Findings

Get detailed PII detection findings for a completed job. Returns the list of PII entities detected in a redaction job, including entity type, detected value, confidence score, and location. Only available after a job has completed. Args: params (GetFindingsInput): Input parameters: - job_id (str): Job ID to retrieve findings for - limit (int): Maximum findings to return (1-500, default 50) Returns: str: JSON with schema: {