pii-detection-pipeline
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data from various sources (CSVs, text files, and S3 buckets) to perform PII detection and redaction. This creates a surface for indirect prompt injection attacks where malicious instructions could be embedded in the processed data to influence an agent's behavior.
- Ingestion points: Data is ingested through
BatchPIIProcessor.process_csv(reading from local paths),BatchPIIProcessor.process_text_files(reading file contents), andMacieIntegration.create_classification_job(scanning S3 buckets). - Boundary markers: The implementation does not include explicit boundary markers or instructions to ignore embedded commands within the data being processed.
- Capability inventory: The skill possesses file system write capabilities (
pd.to_csv,f.write) and network-mediated interaction with AWS services (Macie job creation). - Sanitization: While the skill focuses on redacting PII, it does not sanitize input data for non-PII malicious content like prompt injection sequences.
Audit Metadata