PasteLint cleans the text between writing and publishing

Copy a paragraph from a PDF, put it into a website editor, and it may bring invisible characters, broken line endings and strange spacing along for the ride. PasteLint focuses on this unglamorous stretch between writing and publication. Its maker describes a four-stage workflow: paste, clean, revise and prepare. The product runs in the browser without a login or server-side text-processing service, and its source is public. It is a text editor for making copy usable in another destination, not a place to keep a library of documents.

Why cleaning and rewriting are different jobs

The first stage repairs mechanical damage: hard line breaks from PDFs, zero-width characters, awkward spaces, repeated words and punctuation artifacts. Those transformations should be relatively predictable. PasteLint then offers Second Draft, which applies bounded editorial patterns to tone, length, clarity and structure. The project's documentation stresses that the revised text should preserve actors, conditions, uncertainty and other parts of a sentence's meaning. That is a more modest promise than telling a user that one button can improve any prose. A legal caveat or technical qualifier may sound inelegant and still be essential.

After review, the copy can move into tools for speech and SSML, publishing, email or documentation. Its SSML Builder is useful when text will be read aloud: it prepares markup and chunks for a text-to-speech workflow rather than treating narration as an afterthought. The stages remain separate, so a person can inspect the cleaned version before accepting a stylistic change. There is no automatic handoff that should be mistaken for an editor's final judgment.

A small product with unusually explicit boundaries

PasteLint's repository describes a static HTML, CSS and JavaScript application with no build step and a no-dependency regression suite. The live page is hosted on GitHub Pages. The maker has published an editorial constitution, an architecture account and a journal about observed edge cases. These are not glamorous launch metrics, but they show attention to the failure mode that matters most here: a cleaning tool silently changing what a writer meant. The project says pasted text stays in the browser and is not sent to analytics; users should still inspect the privacy notes and avoid posting sensitive examples to public issue trackers.

The creator has reported using the tool on long IVR and text-to-speech scripts, including scripts that later become speech or published copy. That is a maker-reported use case, not a measured user base or a public performance guarantee. The sources reviewed do not provide revenue, monthly active users or a download total. The visible outcome is a deployed editor, documented workflows and tests that other people can inspect. For a narrow text utility, that evidence is more useful than an invented adoption figure.

Where it fits in a writing workflow

PasteLint is appealing when text has already been written and must survive a change of format: a PDF into a CMS, a draft into narration, or a generated answer into a public page. It is less suited to collaborative drafting, long-form document management or open-ended AI generation. The maker explicitly says it is not a grammar checker, detector-bypass tool or generic writing assistant. That restraint gives the product a clear place in a crowded category. A person still owns the sentence; the software tries to make the transfer and bounded revision legible enough that they can catch a mistake before publication.

Sources: PasteLint live tools; repository, methods and privacy notes.

Project page screenshot captured by CoderHype on October 6, 2026 for this independent editorial listing.