Evenfire is an intelligence layer platform for running LLM agents that take real actions, on your infrastructure
TypeScript
/**
* Provider-fallback (R5) — error classification.
*
* Maps the real `(LlmErrorCode, retryable)` tuples of `core/errors` +
* `errorClassification` onto the closed {@link FailoverClass} catalogue (spec
* §3-R5.2). This is the ONLY place the tuple→class mapping lives, shared by the
* engine's eligibility check and by callers that want to filter classes.
*/
import { LlmErrorCode } from '../../core/errors'
import type { FailoverClass } from './types'
/** The default `triggerOn` set — all four classes (spec §3-R5.2). */
export const ALL_FAILOVER_CLASSES: readonly FailoverClass[] = [
'insufficient_quota',
'auth',
'provider_unavailable',
'rate_limited',
]
/**
* Classify a provider error tuple into a {@link FailoverClass}, or `null` when
* the error is NOT eligible for fallback under any policy.
*
* Deliberate mapping (spec §3-R5.2):
* - `InsufficientQuota` → insufficient_quota (402 / credit balance)
* - `AuthenticationFailed` → auth (401 and 403 collapse)
* - `RateLimited` → rate_limited (429)
* - `ModelOverloaded` → provider_unavailable (5xx / 529)
* - `ControlPlaneUnavailable` → provider_unavailable (a control-plane hop no
* live process answered, #720)
* - `ApiCallFailed ∧ retryable` → provider_unavailable (timeout / network;
* also `classifyUnknown`'s default bucket — restrict `triggerOn` to avoid
* firing on unrecognised errors)
* - `ApiCallFailed ∧ !retryable` → null (400 / validation / content-policy —
* never trigger; masking these would hide bugs)
* - everything else (InvalidResponse, ContextLengthExceeded, ContentFiltered,
* ModelNotAvailable, ToolCallLimitExceeded, StreamDurationExceeded,
* UpstreamRejected) → null
*/
export function classifyFailoverClass(
code: LlmErrorCode,
retryable: boolean
): FailoverClass | null {
switch (code) {
case LlmErrorCode.InsufficientQuota:
return 'insufficient_quota'
case LlmErrorCode.AuthenticationFailed:
return 'auth'
case LlmErrorCode.RateLimited:
return 'rate_limited'
case LlmErrorCode.ModelOverloaded:
case LlmErrorCode.ControlPlaneUnavailable:
return 'provider_unavailable'
case LlmErrorCode.ApiCallFailed:
return retryable ? 'provider_unavailable' : null
default:
return null
}
}