Egonex-AI/Understand-Anything · warning · Error

Freshness request failed

Error message

Freshness request failed

What it means

Thrown by requestFreshnessReport when the HTTP response from the freshness endpoint has a non-ok status. The function fetches the URL with an AbortSignal and no-store cache, then immediately checks response.ok before parsing the body; any 4xx/5xx is fatal and not retried. Callers in startFreshnessRefresh catch it and degrade to an 'unknown/freshness-request-failed' report.

Source

Thrown at understand-anything-plugin/packages/dashboard/src/freshness.ts:173

    !("domain" in value.graphs) ||
    isGraphFreshnessResult(value.graphs.domain)
  );
}

export function shouldRequestFreshness(
  demoMode: boolean,
  demoFreshnessUrl?: string,
): boolean {
  return !demoMode || Boolean(demoFreshnessUrl);
}

export async function requestFreshnessReport(
  url: string,
  signal: AbortSignal,
  fetcher: typeof fetch = fetch,
): Promise<DashboardFreshnessReport> {
  const response = await fetcher(url, { signal, cache: "no-store" });
  if (!response.ok) throw new Error("Freshness request failed");

  const payload: unknown = await response.json();
  if (!isDashboardFreshnessReport(payload)) {
    throw new Error("Freshness response was malformed");
  }
  return payload;
}

interface FreshnessRefreshOptions {
  target: Pick<EventTarget, "addEventListener" | "removeEventListener">;
  load: (signal: AbortSignal) => Promise<DashboardFreshnessReport>;
  onResult: (report: DashboardFreshnessReport) => void;
}

function requestFailedReport(): DashboardFreshnessReport {
  return {
    graphs: {
      knowledge: {

View on GitHub (pinned to 32944829e7)

Solutions

1. Verify the dashboard dev server is running and the freshness URL is correct and reachable.
2. Check the access token used for the freshness endpoint is valid and has not expired.
3. Inspect server logs for the underlying status code returned by the freshness route.
4. Handle the error gracefully — startFreshnessRefresh already maps it to an 'unknown' report, so ensure callers rely on that fallback rather than treating it as fatal.

Example fix

// before
const report = await requestFreshnessReport(url, signal);
// after — tolerate failure and use the degraded report
try { const report = await requestFreshnessReport(url, signal); }
catch { /* fall back to { graphs: { knowledge: { status: 'unknown', reason: 'freshness-request-failed' } } } */ }

// pre-check the endpoint reachability before the freshness call
const probe = await fetch(url, { method: 'HEAD' });
if (!probe.ok) { /* skip freshness, use degraded report */ }

try { const report = await requestFreshnessReport(url, signal); onResult(report); }
catch (e) { onResult({ graphs: { knowledge: { status: 'unknown', reason: 'freshness-request-failed' } } }); }

When it happens

Trigger: The dashboard dev server's freshness endpoint returns 404 (route removed/misconfigured), 401/403 (access token invalid), 500 (server-side git probe failure), or the endpoint is unreachable and a non-ok response is returned by a proxy.

Common situations: Dashboard dev server running a version that lacks the freshness route; access token expired or revoked; reverse proxy returning an error page; the freshness endpoint disabled in a deployment.

Related errors

• Figma API ${path} failed: ${res.status} ${res.statusText}
• Freshness response was malformed
• HTTP ${res.status}
• Source unavailable

AI-assisted analysis of Egonex-AI/Understand-Anything@32944829e7 (2026-08-12). Data as JSON: /api/errors/de64f09a1eae4157. Report an issue: GitHub.