Tier Gating (v0.2.2+)

PayFlux exports support tier-based authority gating. Same data, same pipeline—different explanatory power.

Tiers

| Tier | Label | Purpose | |------|-------|---------| | tier1 | Processor Exposure Monitor | Detection only | | tier2 | Revenue Protection | Interpretation + foresight |

Configuration

PAYFLUX_TIER=tier1   # Default: detection only
PAYFLUX_TIER=tier2   # Adds processor_playbook_context + risk_trajectory

Tier 1 — Detection Only

Tier 1 exports include:

• Risk score, band, and drivers
• Raw metrics and timestamps
• Neutral descriptions of what was detected

Tier 1 exports do not include:

• processor_playbook_context
• risk_trajectory

Tier 2 — Interpretation + Foresight

Tier 2 exports include everything in Tier 1, plus:

1. Processor Playbook Context

Field: processor_playbook_context

Answers: "Why processors care."

Rules:

• Probabilistic language only ("typically," "commonly observed")
• No guarantees
• No prescriptive actions
• No processor-specific rules stated as fact

Example:

"This pattern typically correlates with processor-side monitoring escalation. Rate limiting or velocity checks are commonly triggered at this threshold."

2. Risk Trajectory

Field: risk_trajectory

Answers: "Is this getting worse—and how fast?"

Shows:

• Directional trend (accelerating / stable / decelerating)
• Magnitude relative to baseline (e.g., "3× above baseline")
• Time-based framing

Example:

"Pattern accelerating: ~3.2× above baseline over the last 5 minutes."

Export Examples

Tier 1 Export

{
  "event_id": "550e8400-e29b-41d4-a716-446655440000",
  "event_type": "payment_failed",
  "processor": "stripe",
  "processor_risk_score": 0.72,
  "processor_risk_band": "high",
  "processor_risk_drivers": ["high_failure_rate", "retry_pressure_spike"]
}

Tier 2 Export

{
  "event_id": "550e8400-e29b-41d4-a716-446655440000",
  "event_type": "payment_failed",
  "processor": "stripe",
  "processor_risk_score": 0.72,
  "processor_risk_band": "high",
  "processor_risk_drivers": ["high_failure_rate", "retry_pressure_spike"],
  "processor_playbook_context": "This pattern typically correlates with processor-side monitoring escalation. Rate limiting or velocity checks are commonly triggered at this threshold. Elevated failure rates are commonly interpreted as degraded transaction quality. Retry clustering typically signals infrastructure stress or integration issues to processor risk systems.",
  "risk_trajectory": "Pattern accelerating: ~3.2× above baseline over the last 5 minutes."
}

Guardrails (Non-Negotiable)

• PayFlux does not recommend actions
• PayFlux does not claim insider processor rules
• PayFlux does not make decisions
• Everything is framed as observed behavior + momentum