What a rule can prove, and what it cannot
A rule can prove that you typed "the" twice in a row. It can prove that "everyone are waiting" disagrees with itself, and that "could of" is not standard English. These are mechanical facts, checkable with a pattern and nothing else.
A rule can never prove that a sentence is any good, so this checker does not try. It runs fifteen patterns over your English text and reports what each one finds, with the offending words, a short message, and a correction where a correction is unambiguous. It does not know what your sentences mean, which is why the panel says its suggestions can be wrong in context. Read that as a description of the tool, not as legal cover. The patterns it knows are the mistakes the eye repairs on its own, which is why they survive a normal read-through.
The fifteen rules this checker ships
There are no options to configure. One text field, and every rule runs on every pass.
| Rule | Catches | Suggestion |
|---|---|---|
| Repeated word | "the the draft" | removes the duplicate |
| Lowercase pronoun | "i could help" | "I" |
| Modal plus "of" | "could of finished" | "could have" |
| Contraction check | "your welcome" | "you're welcome" |
| Singular "don't" | "it don't matter" | "it doesn't matter" |
| Plural with "was" | "They was ready" | "They were ready" |
| Singular subject with "are" | "Everyone are waiting" | "Everyone is waiting" |
| Pronoun case after "between" | "between you and I" | "between you and me" |
| Wordiness | "in order to finish" | "to finish" |
| Hyphen after an -ly word | "quickly-but" | none offered |
| Countable "less" | "less people" | "fewer people" |
| Nonstandard word | "irregardless" | "regardless" |
| "its" before an article | "Its a shame" | "it's a shame" |
| Missing space after , ; : | "first,second" | "first, second" |
| Space before , . ? | "fine ." | "fine." |
The rows are narrower than their names suggest, and the width matters when you read a clean report. The "less" rule covers only four nouns: people, items, words, and books. "Less cars" and "less water" are not flagged, even though the first is the error the rule exists for. The missing-space rule fires after commas, semicolons, and colons, never after a period, so "One sentence.Next one." passes untouched.
Each finding comes with its character position. Only the repeated-word rule matches across a line break, so a word left at the end of one line and repeated at the start of the next is still caught. The hyphen rule is the opposite case: it needs both halves of the hyphenated word on the same line, so a family-owned split at the line ending escapes it entirely and the pair is never compared.
Where the rules get it wrong
Every one of these is a real output from the tool above.
"your right" is not always wrong. The contraction rule fires on "your welcome", "your right", and "your wrong" without checking what follows. Type "You have your right to vote" and it suggests "you're right", turning a correct sentence into an incorrect one.
"could of course" is not a misspelling. The modal rule matches "could of" wherever the two words appear. Write "We could of course wait" and it offers "could have", producing "We could have course". Applying it introduces an error while fixing nothing.
"i.e." is a Latin abbreviation. The lowercase pronoun rule matches any standalone "i", and the surrounding periods do not protect it. "Use the short form, i.e. this one" gets flagged, and applying the suggestion yields "I.e."
"family-owned" is not an adverb. The hyphen rule looks for a hyphenated compound whose first half ends in the letters l-y, which catches "family-owned", "assembly-line", and "Italy-based". It offers no replacement, so it is harmless, though it will sit in your list.
"in order to" is a style preference. The tool suggests "to" every time. Sometimes the longer phrase earns its extra syllables, particularly where a bare "to" would read as a preposition. It is the clearest case for declining a suggestion, and here you can only skip it.
The hardest case in English grammar is not in the set
You may have noticed that "a hour" and "an university" produce nothing. There is no article rule in the set, and the reason is worth understanding. The choice between "a" and "an" depends on the sound of the next word, not its spelling. "Hour" begins with a consonant letter and a vowel sound. "University" begins with a vowel letter and a consonant sound, "yoo". A pattern that reads letters cannot see either fact.
A rule could guess, and the guesses would fail on exactly the words that matter: hour, honest, university, unique, European, one. So this checker stays out of it, and the article problem stays one of the few decisions you make with your ear. Read the sentence aloud. If the article and the next word collide in your mouth, change the article.
Who gets real value from a mechanical pass
People editing their own work. The repeated word is the gift of this rule set, because it is the error you cannot see. When you revise a sentence in place, you often leave the old word behind, and "the the" or "was was" reads as a stutter that your brain silently repairs. No amount of re-reading catches it reliably. A pattern match catches it every time.
Writers whose first language is not English. Several rules target errors that transfer from other languages: the doubled word, the modal plus "of", the "everyone are" agreement slip, the "between you and I" habit. Concrete corrections are more useful than a general note that the grammar needs work.
Anyone proofreading a document they did not write. A quick pass turns up the mechanical slips in a transcript or a client draft before you spend attention on the argument.
Limitations, stated plainly
The checker cannot judge meaning, so it cannot tell you that a sentence says the opposite of what you intended, that a pronoun has no clear antecedent, or that a paragraph contradicts the one above it. It has no idea whether your facts are right or whether your tone fits your audience.
It will not catch a comma splice, a dangling modifier, a tense shift, or a missing comma. It does not check spelling beyond the handful of words built into its rules.
A clean report is not a guarantee of anything except that fifteen patterns did not match. Read it as a floor, not a ceiling. The article on grammar mistakes that make writing look amateur covers the errors that live above this checker's reach.
Frequently Asked Questions
Why didn't it flag "a hour" or "an university"?
Because the checker has no article rule. Choosing between "a" and "an" depends on the sound of the following word rather than its letters, and a pattern that reads letters would guess wrong on hour, honest, university, unique, and one. That decision stays with your ear.
Why does it want to change "in order to" to "to"?
That rule is a style preference, not an error. The tool offers the shorter form because edited prose usually prefers it. If the longer phrase reads better in your sentence, skip the suggestion. There is no Ignore button on this panel, so it will come back on the next pass.
It flagged "your right to vote" as "you're right". Am I wrong?
No, the checker is. The contraction rule fires on "your welcome", "your right", and "your wrong" without looking at what follows, and in "your right to vote" the word "right" is a noun rather than the start of a contraction. Decline the suggestion.
Does a clean report mean the document is error-free?
No. It means fifteen specific patterns did not match. Comma splices, tense shifts, unclear pronouns, and every error of meaning fall outside the rule set entirely. Treat the report as a floor under your own reading, never as a replacement for it.
Why is there no Ignore button on this checker?
Because the two panels are built differently. The spelling checker keeps a list of words you have exempted, and this one has no such memory, since a pattern can match many different sentences. The cost is that style suggestions reappear on every pass.
Run it again after you edit in place
Here is the one habit worth taking away. Any time you revise a sentence by deleting a word and typing a replacement, run the check once more before you move on. That is the moment "the the", "was was", and "to to" appear, because the old word survives the edit and your eye does not notice. Ten seconds, one rule, and the error you are least likely to find by reading.