ggithub.com

The file contains no "Sitemap:" line.

Consequence. Crawlers have to discover every URL by following links. Pages that are deep, newly published, or poorly linked internally take longer to be found, and some may never be. A sitemap is the most direct way to hand a crawler the full list.

If that is not intended. Publish an XML sitemap and declare it with an absolute URL, e.g. "Sitemap: https://example.com/sitemap.xml". The line is global — it belongs outside any User-agent group and applies to every crawler.

Lines that do not parse as a known directive: invalid_path.

Consequence. A crawler skips what it cannot parse, without reporting anything. A misspelled "Dissalow" leaves a path fully crawlable while the file appears to restrict it.

If that is not intended. Correct the spelling of each field name, and make sure every line follows "Field: value". Rules must also appear after a User-agent line to belong to a group.

• Path '*/.git/' should start with '/'L51invalid_path
• Path '*/tarball/' should start with '/'L58invalid_path
• Path '*/zipball/' should start with '/'L59invalid_path
• Path '*/.git/' should start with '/'L237invalid_path
• Path '*/tarball/' should start with '/'L244invalid_path
• Path '*/zipball/' should start with '/'L245invalid_path
• Path '*/tarball/' should start with '/'L305invalid_path
• Path '*/zipball/' should start with '/'L306invalid_path
• Path '*/tarball/' should start with '/'L315invalid_path
• Path '*/zipball/' should start with '/'L316invalid_path
• Path '*/tarball/' should start with '/'L325invalid_path
• Path '*/zipball/' should start with '/'L326invalid_path
• Path '*/.git/' should start with '/'L369invalid_path
• Path '*/tarball/' should start with '/'L376invalid_path
• Path '*/zipball/' should start with '/'L377invalid_path

Rules name version-control directories, credential files, databases, dumps, archives or logs.

Consequence. If this file is served, it is a critical exposure; listing it here suggests it exists. robots.txt is public and among the first files a scanner reads, and it protects nothing: it only tells readers where to look.

If that is not intended. Check whether these paths answer. Remove the files from the web root or block them at the server, then delete the lines. If a .env file or a .git directory was ever served, rotate every credential it held.

• Disallow: /.git/L50Secret file or directory (.git)
• Disallow: /.git/L236Secret file or directory (.git)
• Disallow: /.git/L368Secret file or directory (.git)

Crawl-delay was never part of the standard. Google has never supported it; Bing and Yandex do.

Consequence. Any rate limit intended for Googlebot is not being applied. If the goal was to reduce load from Google specifically, it is not working.

If that is not intended. Set the crawl rate for Google in Search Console instead. Keep the directive if the limit is meant for Bing or Yandex, which do honour it.

• Crawl-delay: 1L8
• Crawl-delay: 1L332

A broader rule of the same type already covers these paths.

Consequence. These lines change nothing. They are harmless, but they add length to a file that has a size limit, and they make the real policy harder to read.

If that is not intended. Remove the redundant rules, or narrow the broader rule if it is catching more than intended.

• allow: /marketplace/actions/download*L23Already covered by "allow: /marketplace" on line 14
• disallow: /copilot/c/L65Already covered by "disallow: /copilot/" on line 64
• disallow: /copilot/c/L251Already covered by "disallow: /copilot/" on line 250
• disallow: /copilot/c/L309Already covered by "disallow: /copilot/" on line 308
• disallow: /copilot/c/L319Already covered by "disallow: /copilot/" on line 318
• disallow: /copilot/c/L329Already covered by "disallow: /copilot/" on line 328
• disallow: /copilot/c/L383Already covered by "disallow: /copilot/" on line 382

Blocked: Bytespider.

• BytespiderByteDanceL206

  ByteDance model training

  "Disallow: /" blocks Bytespider.

If that is the intention. If your intention is to prevent this content being used as model training data, the current configuration achieves that for these crawlers — the rest of the group is still allowed.

Consequence. These crawlers cannot retrieve content from the site, so it will not be collected as training data. They send no traffic back, and blocking them does not affect search rankings.

If that is not intended. Remove the site-wide block, or replace it with narrower path restrictions.

Wildcard patterns name no location, but suggest such files may exist under public paths.

Consequence. A pattern such as "/*.sql$" discloses nothing a reader can request directly. It does say what kind of file the site expects to have lying around.

If that is not intended. Verify that files matching these patterns are not served.

• Disallow: */.git/L51Secret file or directory (.git)
• Disallow: */.git/L237Secret file or directory (.git)
• Disallow: */.git/L369Secret file or directory (.git)

All 6 checked search engines are allowed at the site root.

All 4 checked AI answer engines are allowed at the site root.

All 3 checked user-triggered AI fetches are allowed at the site root.

All 5 checked social link previews are allowed at the site root.

The page answered HTTP , so its meta robots tags and X-Robots-Tag headers were not read.

Consequence. Nothing is known about whether this page asks to be kept out of search results. Its absence of directives here proves nothing either way.

If that is not intended. Analyse the page again once it answers normally to our bot.

Comfortably within the 500 KB that crawlers will read.