jackwener/OpenCLI · error · CommandExecutionError

captured.error

Error message

captured.error

What it means

In the segments command, after the player-extraction path fails, the CLI falls back to extractSegmentsFromNetworkCapture over the browser's network capture. If that function returns {error}, the error is thrown as a CommandExecutionError. The capture path reports an error when captured timedtext entries exist but none could be parsed into segments.

Source

Thrown at clis/youtube/transcript.js:384

            const parsed = parseJson3(text);
            if (parsed.error) return { error: parsed.error };
            if (parsed.rows.length > 0) return parsed.rows;
          }

          return null;
        } finally {
          try { player?.pauseVideo?.(); } catch {}
          if (originalFetch) globalThis.fetch = originalFetch;
          if (OriginalXHR) globalThis.XMLHttpRequest = OriginalXHR;
        }
      })()
    `);
        let segments = normalizeSegmentsPayload(playerResult, 'player caption extraction', { allowNull: true });
        if (!segments && canCapture) {
            try {
                const captured = extractSegmentsFromNetworkCapture(await page.readNetworkCapture(), lang, videoId);
                if (captured.error) {
                    throw new CommandExecutionError(captured.error);
                }
                if (captured.segments.length > 0) {
                    segments = captured.segments;
                }
            }
            catch (err) {
                if (err instanceof CommandExecutionError)
                    throw err;
                // Keep existing fallback path when capture is unavailable.
            }
        }
        if (!segments) {
            await prepareYoutubeApiPage(page);
        }
        // Fallback: get caption track URL from watch page HTML
        const captionData = segments ? null : unwrapBrowserResult(await page.evaluate(`
      (async () => {
        const extractJsonAssignmentFromHtml = ${extractJsonAssignmentFromHtml.toString()};

View on GitHub (pinned to 49907e53dc)

Solutions

1. Restart the browser session to clear stale network captures, then reload the video page so fresh timedtext requests are captured.
2. Check the embedded captured.error message for the underlying parse failure cause.
3. Ensure consent/bot checks passed in the automated tab before the timedtext request fires.
4. Update the CLI if YouTube changed the timedtext response format or pot token requirements.

Example fix

// before
const captured = extractSegmentsFromNetworkCapture(await page.readNetworkCapture(), lang, videoId);
if (captured.error) throw new CommandExecutionError(captured.error);
// after
const captured = extractSegmentsFromNetworkCapture(await page.readNetworkCapture(), lang, videoId);
if (captured.error) {
  console.error('network capture fallback failed:', captured.error);
  // trigger a fresh page reload to recapture timedtext
}

const capture = await page.readNetworkCapture();
const hasTimedtext = Array.isArray(capture) && capture.some(e => String(e?.url || '').includes('/api/timedtext'));
if (!hasTimedtext) {
  // nothing captured yet; reload the page before using the fallback path
}

function isCaptureResult(v) {
  return !!v && typeof v === 'object' && (Array.isArray(v.segments) || typeof v.error === 'string');
}

try {
  const captured = extractSegmentsFromNetworkCapture(await page.readNetworkCapture(), lang, videoId);
  if (captured.error) throw new Error(captured.error);
} catch (err) {
  console.error('capture fallback failed:', err.message);
  // reload page and recapture fresh timedtext responses
}

When it happens

Trigger: `segments` -> fallback path where page.readNetworkCapture() contains timedtext entries matching lang/video, but parseJson3Segments fails on every responsePreview, so extractSegmentsFromNetworkCapture returns {error: '...'} which is rethrown.

Common situations: Stale signed timedtext URLs in the daemon-shared tab returning error bodies; YouTube serving bot-check HTML in captured timedtext responses; SPA navigation leaving captures from prior videos that pass URL filtering but have expired pot tokens.

Related errors

• Gmail ${operation} requires browser response interception
• Gmail ${operation} could not start browser response intercep
• Gmail ${operation} capture included a response without a bod
• Gmail ${operation} capture
• Unexpected YouTube probe: ${JSON.stringify(probe)}

AI-assisted analysis of jackwener/OpenCLI@49907e53dc (2026-08-29). Data as JSON: /api/errors/3cfb60e950862cb9. Report an issue: GitHub.