Hotel Chain
Redact guest and payment data, block risky exports, screen configured terms, check output quality and retain hotel records.
What it checks
| Policy | On | Checks | Action |
|---|---|---|---|
| prompt-injection | Request | Detects jailbreak and prompt-injection signals | Block |
| pii-detector | Request | 1 custom identifier pattern | Redact |
| dlp-filter | Request | 3 blocked terms such as “private key” and “bulk customer export” | Block |
| safety-filter | Request | Blocks or escalates configured dangerous terms | Block |
| quality-scorer | Response | Accuracy ≥ 0.80; relevancy ≥ 0.78; combined score ≥ 0.78; at least 120 characters and 2 sentences | Block |
| audit-logger | Request | Adds an allow-only audit marker to the decision | Record |
Policy YAML
pack:
id: "cfg_hotel_chain"
name: "hotel-chain"
version: "0.1.0"
enabled: true
author: "verdictan"
description: "Hotel controls for guest privacy, payment-token data, default safety keyword filtering, and records."
policies:
chain:
- "prompt-injection"
- "pii-detector"
- "dlp-filter"
- "safety-filter"
- "quality-scorer"
- "audit-logger"
policy:
prompt-injection: {}
pii-detector:
action: redact
detect_patterns:
- "(?i)(?:payment|card)\\s+token\\s*[:#=]?\\s*[A-Za-z0-9_\\-]{8,64}"
dlp-filter:
blocked_terms:
- "private key"
- "bulk customer export"
- "gift card pin"
action: block
safety-filter: {}
quality-scorer:
min_output_chars: 120
min_sentences: 2
thresholds:
min_aggregate: 0.78
min_relevancy: 0.78
min_accuracy: 0.80
weights:
relevancy: 0.5
accuracy: 0.5
audit-logger: {}