jackwener/OpenCLI · error · ArgumentError

--content exceeds ${MAX_TEXT_SOURCE_BYTES} bytes; split into

Error message

--content exceeds ${MAX_TEXT_SOURCE_BYTES} bytes; split into smaller sources or upload as a file.

What it means

parseSourceText throws this ArgumentError when the --content string's length exceeds MAX_TEXT_SOURCE_BYTES, the library's cap for inline text sources. The message advises splitting the text into smaller sources or uploading it as a file instead.

Source

Thrown at clis/notebooklm/add-source.js:86

    if (!url) return '';
    let parsed;
    try {
        parsed = new URL(url);
    } catch {
        throw new ArgumentError(`Invalid source URL: "${url}"`, 'URL must be a valid http:// or https:// URL.');
    }
    if ((parsed.protocol !== 'http:' && parsed.protocol !== 'https:') || !parsed.hostname) {
        throw new ArgumentError(`Invalid source URL: "${url}"`, 'URL must start with http:// or https://.');
    }
    return parsed.toString();
}

export function parseSourceText(value) {
    if (value === undefined || value === null) return null;
    const text = String(value);
    if (!text.trim()) throw new ArgumentError('--content must not be empty');
    if (text.length > MAX_TEXT_SOURCE_BYTES) {
        throw new ArgumentError(`--content exceeds ${MAX_TEXT_SOURCE_BYTES} bytes; split into smaller sources or upload as a file.`);
    }
    return text;
}

export function parseSourceTitle(value, fallback) {
    const title = String(value ?? '').trim();
    return title || fallback;
}

export function buildAddSourceFromUrlArgs(projectId, url) {
    return [[[null, null, [url]]], projectId];
}

export function buildAddSourceFromTextArgs(projectId, title, content) {
    return [[[null, [title, content], null, 2]], projectId];
}

function toExcludedUuidSet(excludedIds) {

View on GitHub (pinned to 49907e53dc)

Solutions

Split the text into multiple chunks under the limit and add each as a separate source.
Write the text to a file and use --file instead (a different, typically larger limit may apply).
Trim the content to the relevant excerpt before passing it.
Split on paragraph/heading boundaries with a script to keep chunks semantically coherent.

Example fix

// before
addSource({ content: entireBook });
// after
const chunks = splitByParagraphs(entireBook, MAX_TEXT_SOURCE_BYTES);
for (const c of chunks) addSource({ content: c });

Defensive patterns

Strategy: validation

Validate before calling

if (text.length > MAX_TEXT_SOURCE_BYTES) {
  throw new Error(`Content too large inline: ${text.length} > ${MAX_TEXT_SOURCE_BYTES}; split or use --file`);
}

Type guard

function fitsInlineText(v, max = MAX_TEXT_SOURCE_BYTES) { return typeof v === 'string' && v.length <= max && v.trim().length > 0; }

Try / catch

try {
  await addSource({ content: text });
} catch (e) {
  if (e instanceof ArgumentError && e.message.includes('--content exceeds')) {
    for (const chunk of splitByParagraphs(text, MAX_TEXT_SOURCE_BYTES)) {
      await addSource({ content: chunk });
    }
  } else throw e;
}

Prevention

Measure text.length against MAX_TEXT_SOURCE_BYTES before calling.
Split long documents on paragraph/heading boundaries.
Prefer --file for large documents instead of inline --content.
For multi-byte text, remember real byte size can exceed the char count.

When it happens

Trigger: Passing a very long document (book chapter, full transcript, large log dump) as --content whose String length exceeds MAX_TEXT_SOURCE_BYTES; note the check uses text.length (UTF-16 code units) as a byte proxy.

Common situations: Pasting entire meeting transcripts or books; piping a large file's contents into --content instead of using --file; multi-byte content whose real byte size is larger than the char count suggests.

Related errors

AI-assisted analysis of jackwener/OpenCLI@49907e53dc (2026-08-29). Data as JSON: /api/errors/92f80847e859fc82. Report an issue: GitHub.