Last updated: October 2026
About This Site
Barber Sped Hub is a free, independent personal project built by a Texas educational diagnostician. It is not a vendor or commercial product, and it is not affiliated with any school district, education service center, test publisher, or the Texas Education Agency.
These services are involved in running the site:
- Netlify hosts the site's pages.
- Railway runs the hub's small server, which accepts AI requests only from barbersped.org, passes them to the AI service, and returns the draft. It has no database and does not log what it receives.
- Anthropic (Claude API) generates the AI drafts.
- Google Analytics counts page visits; it does not read what you type. Microsoft Clarity measures how people use the hub's reference pages and does not run at all on pages that handle student information — the Report Writer, REED Planner, FIE Addendum, What Does This Subtest Measure?, Score Pattern Decoder, Error Analyzer, and TEKS CBM tools, and every form or tool where you type or upload information about a student.
- Formspree delivers messages sent through the contact form.
On pages that handle student information, the code libraries that read PDF and Word files are served from barbersped.org itself, and a browser security policy allows those pages to send data only to the hub's own server (plus page-visit counts to Google Analytics). Any other destination is blocked by the browser.
What the AI Report Writer Sends
When you ask the Report Writer to draft a section, it sends what that section needs over an encrypted (HTTPS) connection. Registered names are replaced before anything is sent.
| Data Type | Sent? | Notes |
|---|---|---|
| Assessment scores, eligibility area selections, evaluation type | Yes | Needed to draft the narratives |
| Grade, age in years and months, pronouns | Yes | Used for grammar and age-appropriate wording; no birthdate is collected |
| Text you type or paste (observations, teacher and parent input, MTSS, interview notes) | Yes, if entered | Registered names are replaced, and email addresses, phone numbers, ID numbers (six or more digits), and dates following a birthdate label are removed automatically. Other dates, such as screening dates, stay because the report needs them. Leave out addresses and any name you have not registered; no scrubber can catch every identifier in free text. |
| Uploaded teacher, parent, MTSS, and REED documents | Text only | Read in your browser; labeled names, registered names, and the identifiers above are removed before the text is placed in the form. The file is never uploaded. |
| Uploaded score-report PDFs | Text only | Read in your browser; registered names, labeled name/ID/school fields, all dates, and long ID numbers are removed, and only that text is sent to read the scores. The file is never uploaded. |
| Photos of student work | Yes, as images | Images cannot be scrubbed. Cover or crop out names and dates before taking the photo. |
| Student name, evaluator name, and other names (scrub fields) | Never | Used only in your browser to find and replace names; not saved |
Other AI Tools on the Hub
- REED Planner: uses the same scrub fields and scrubber as the Report Writer. Uploaded PDF and Word files are read in your browser, and the file is never uploaded. Registered names, labeled names, email addresses, phone numbers, ID numbers (six or more digits), and birthdates are removed from the extracted text, and you can review and edit that cleaned text before drafting; only it is sent. Notes and other text you type or paste are scrubbed the same way before sending. Scanned or image-only PDFs are not accepted, because names cannot be removed from them; paste the sections you need into the text box instead, and pages of a PDF that look scanned are left out. The scrub fields, uploaded text, and pasted text are kept in the page's memory only: they are not saved, not included in Save Session, and are empty again after a refresh.
- Score Pattern Decoder: PDF score reports are read in your browser and sent as text with labeled name, ID, birthdate, and school fields, all dates, and long ID numbers removed. This page has no name field, so a name printed without a label stays in the text. Screenshots and photos are sent as images, so crop names out first.
- Error Analyzer: photos of student work are sent as images. Crop names out first.
- FIE Addendum and TEKS CBM tools: send only the text you enter.
Data Storage & Retention
Nothing you enter is stored by the hub. There is no database, login, or user account, and the hub's server does not keep or log the requests it passes along.
Your own browser does keep a copy. The Report Writer autosaves your work in this browser on this device so a refresh doesn't lose it, and deletes that autosave automatically after 14 days without use. The name-scrub fields are not saved. Clear All, or "Clear saved session," erases the autosave. Save Session downloads a file to your computer; it is never sent to the site. On a shared computer, clear your session when you finish.
The contact form sends only the name, email address, and message you enter, through Formspree. It is used only to reply to you.
FERPA and District Policy
The Family Educational Rights and Privacy Act (FERPA, 20 U.S.C. § 1232g) protects the privacy of student education records. The hub is designed to minimize what leaves your device, but it is a personal project, not a vendor, and there is no signed student data privacy agreement in place with any district.
If your district requires a data privacy agreement, or approval through the Texas Student Privacy Alliance, before staff use an online tool, check with your technology or data privacy office before using the AI features.
Users remain responsible for:
- Listing parent, sibling, teacher, and campus names in the Other Names field when they appear in what you enter or upload
- Leaving ID numbers, birthdates, addresses, and other people's names out of text fields, and removing them from documents and photos the hub sends whole
- Following their district's data governance and technology use policies
- Reviewing all drafted text and finalizing it within secure, district-approved systems before it becomes part of an official record
This page is not an agreement or a guarantee of FERPA compliance on behalf of any district.
Anthropic API
AI drafting is powered by the Anthropic Claude API under Anthropic's commercial terms. Data sent through the API is governed by Anthropic's Privacy Policy and commercial terms, which state that API data is not used to train Anthropic's models. Anthropic automatically deletes API inputs and outputs within 30 days, except where it must keep them to enforce its usage policy or by law (Anthropic's retention policy). Policies can change, so review Anthropic's current terms independently.
Questions or Concerns
If you have questions about data practices on this hub, use the feedback form on the main page. For questions about your district's data governance policies, contact your district's data privacy officer or technology department.