agentfootprint
Version:
The explainable agent framework — backtrack a wrong answer to the exact context that caused it (evidence, not guesses). Built on footprintjs.
35 lines • 1.45 kB
JavaScript
/**
* statedWait — how long a failure SAID to wait before asking again.
*
* Pattern: pure reader (no I/O). Role: the ONE place the resilience layer
* learns a server's stated wait.
*
* An error declares it; this module never parses prose or headers. The adapter
* that knows its wire does that (a `Retry-After` header, a gateway's "Try again
* in 4 seconds") and puts the answer on the error:
*
* - `retryAfterMs` — the declared field, in milliseconds. Every adapter in
* this library that can learn a wait sets it.
* - `retryAfterSeconds` — read as a fallback, because errors built outside
* this library (and `InvokeModelGatewayError`'s header-only field, which
* predates `retryAfterMs`) carry the wait in seconds.
*
* A value that is not a finite, non-negative number is treated as absent, so a
* garbled field degrades to the policy's own schedule rather than to a wrong
* wait.
*/
/** The wait a failure stated, in ms — or `undefined` when it stated none. */
export function statedRetryAfterMs(err) {
if (err === null || typeof err !== 'object')
return undefined;
const e = err;
if (isWait(e.retryAfterMs))
return e.retryAfterMs;
if (isWait(e.retryAfterSeconds))
return e.retryAfterSeconds * 1000;
return undefined;
}
function isWait(value) {
return typeof value === 'number' && Number.isFinite(value) && value >= 0;
}
//# sourceMappingURL=statedWait.js.map