FDA 21 CFR Part 11

Map FDA 21 CFR Part 11 controls for validation, audit trails, access, retention and electronic signatures.

Control framework · Regulated SaaS, Clinical zero-retention, Private cloud

What it checks

Policies in the FDA 21 CFR Part 11 template, in the order the gateway runs them
PolicyOnChecksAction
prompt-injectionRequestDetects jailbreak and prompt-injection signalsBlock
rbacRequestRequires the X-User-ID, X-Org-ID and X-Study-ID headersBlock
pii-detectorRequestHealthcare identifiers; 3 custom identifier patternsRedact
dlp-filterRequest4 custom patternsBlock
citation-verifierResponseRequires cited sources; confidence ≥ 0.90Flag
quality-scorerResponseAccuracy ≥ 0.90; combined score ≥ 0.85Block
human-oversightResponseReturns an escalated result instead of the assistant contentEscalate
audit-loggerRequestAdds an allow-only audit marker to the decisionRecord
Policy YAML
pack:
  id: "fw_pharma_baseline"
  name: "pharma-baseline"
  version: "0.1.0"
  enabled: true
  author: "verdictan"
  description: "Life-sciences baseline starter for regulated records, sources, and supervised release."
policies:
  chain:
    - "prompt-injection"
    - "rbac"
    - "pii-detector"
    - "dlp-filter"
    - "citation-verifier"
    - "quality-scorer"
    - "human-oversight"
    - "audit-logger"
policy:
  prompt-injection: {}
  rbac:
    deny_if_missing: ["X-User-ID", "X-Org-ID", "X-Study-ID"]
  pii-detector:
    healthcare_mode: true
    action: redact
    detect_patterns: ["(?i)(?:patient\\s*id|patient\\s*#|pt\\s*id)\\s*[:#-]?\\s*[A-Z0-9-]{6,16}", "(?i)(?:subject|participant)\\s*(?:id|#|number)\\s*[:#-]?\\s*[A-Z0-9-]{4,20}", "(?i)\\binvestigator\\s*(?:id|#|number)\\s*[:#-]?\\s*[A-Z0-9-]{4,20}"]
  dlp-filter:
    detect_patterns: ["(?i)\\bclinical\\s+trial\\s+(?:data|record|results)\\b", "(?i)\\badverse\\s+event\\s*(?:report|id|#)?\\s*[:#-]?\\s*[A-Z0-9-]{4,20}", "(?i)\\bbatch\\s+(?:record|lot|#)\\s*[:#-]?\\s*[A-Z0-9-]{4,20}", "(?i)\\bdeviation\\s+(?:report|record|notice)\\b"]
    action: block
  citation-verifier:
    require_sources: true
    min_confidence: 0.90
  quality-scorer:
    thresholds:
      min_aggregate: 0.85
      min_accuracy: 0.90
  human-oversight:
    action: escalate
  audit-logger: {}