What the parser reports
Directives are grouped by User-agent, with the Allow and Disallow rules that belong to each one plus any declared Sitemap lines. That makes it obvious when a rule written for one specific bot is actually applying to every crawler, which is the costliest and most common mistake.
The mistake that costs the most traffic
A Disallow: / under User-agent: * blocks crawling of the entire site. It usually arrives by being promoted from a staging environment, where that exact rule is correct. Nothing alerts you: indexing simply decays until pages start dropping out of results.
Blocking a crawl is not deindexing
Robots.txt controls whether a crawler may fetch a URL, not whether it may list it. A page blocked by Disallow can still appear in Google if external links point to it, because the engine knows the URL exists even though it cannot read the content. Removing a page from the index requires a noindex tag, and that means the crawler must be allowed to read the page in the first place.