fullrefit.io The fullrefit.io logo, a red square with a white f slash mark followed by the fullrefit.io wordmark. f/ fullrefit.io
Everything inside

One small service for the part most pages get wrong.

fullrefit.io is designed around the moment an AI agent fetches a document. The feature set stays close to that job.

01 · Direct reading

The public route returns the complete Markdown document directly with HTTP 200 and a plain-text content type.

No login, JavaScript shell, or URL-shortener redirect is required for public reads.

02 · Short branded routes

Publish a document at a clean slug such as /wrap instead of handing a reader a long raw storage URL.

The branded route is the human-facing entry point. The response body remains machine-facing text.

03 · Revision history

Authenticated updates increment the document revision and preserve the previous version in D1.

Use revision numbers when a document changes over time and hashes when the body must be compared exactly.

04 · Hash verification

The service computes SHA-256 for stored Markdown and the publisher verifies the public response after writing.

A successful API response is not treated as proof until the public body matches the source.

05 · Browser authoring

Write, publish, and update documents from a signed-in studio on this site. No terminal, no local setup, and no credential pasted into a page.

Writing is protected. Reading is not. The write credential stays on the server and never reaches the browser.

06 · Human-facing context

The service has a normal landing page, feature page, and how-to guide alongside the raw document routes.

People get explanation and orientation. Agents get the exact Markdown body.

What it does not promise.

The service is intentionally narrow. It does not pretend that a clean response solves authentication, source quality, or deployment verification by itself.

No invented content

fullrefit.io serves what you publish. It does not rewrite a document to make it sound better.

No public write access

Public readers can fetch published documents. They cannot change them, and they never see the studio.

No browser extension yet

A Chrome Extension is a separate product surface with its own identity, permissions, and store requirements.