A rule set is an ordered list of your verified gateway connections. Each rule holds one priority slot in the failover chain: 
{
  "priority": 1,
  "gatewayConfigId": "<connection id>",
  "failoverTriggers": {
    "hardDeclines": true,
    "timeouts": true,
    "gatewayErrors": true
  },
  "timeoutMs": 1800,
  "enabled": true
}
SettingMeaning
priority1 is attempted first; on a retryable failure the engine moves to the next enabled priority
failoverTriggers.hardDeclinesFail over on retryable card declines (do_not_honor, insufficient_funds, …)
failoverTriggers.timeoutsFail over when the gateway exceeds the per-attempt timeout
failoverTriggers.gatewayErrorsFail over on gateway 5xx / infrastructure errors
timeoutMsPer-attempt timeout for this gateway
enabledToggle a rule without recreating the set
Fraud-class declines (stolen/lost card, fraud suspected, invalid card) are never failed over, regardless of your trigger settings. Re-presenting a card the issuer flagged as fraudulent damages your standing with issuers — SwitchFlow excludes these by construction. See Failover behavior.

​
Versioning

Saving rules always creates a new version and activates it atomically — in-flight payments finish on the version they started with. Previous versions are kept for audit and are visible in the dashboard. Toggling a single rule’s enabled flag is the one quick action that does not create a new version.

​
Recommendations

  • Keep at least two verified gateways in every rule set — with one gateway there is nothing to fail over to.
  • Use your historically better-performing gateway (see Analytics → auth rate by gateway) as priority 1, and revisit the order monthly.
  • Leave all three failover triggers on unless you have a specific reason not to; the fraud-class exclusion already protects you from harmful retries.