SENTINEL·IEP · INFORMATIONAL TOOL ONLY · NOT LEGAL ADVICE

Community Reports

Anonymized accounts from other subscribers about how their districts have behaved in IEP meetings.

The Community Reports feature lets subscribers share what happened in their districts — what was said, what was refused, what commitments were made — and read what's happened to others. Identifying information is stripped automatically, and submissions go through two-model AI moderation for defamation risk before anything is published.

How to open it

Click Reference in the top nav, then the Community Reports tab.

What you'll see

Each report is shown with:

You will never see the specific state, district name, submitter name, submitter license key, or any other identifying details. Those are stored on the admin side only.

Submitting a report

Click Share what happened (the purple button in the top right of the Reference dialog) to open the submission form.

1 Pick your state. This selects the right defamation notice for you to acknowledge. Your state is never shown to other subscribers — only the broader region (e.g. "Midwest").
2 Pick your district size. In states with only one very large district (e.g. South Dakota), the app warns you that picking "Large" can identify your district even though we won't name it. Pick "Medium" instead or skip the size if you're worried.
3 Optionally name your district (admin-only). If you provide a district name, only the admin (for moderation purposes) can see it. Subscribers cannot. The district name is also explicitly replaced with [district] in the scrubbed narrative.
4 Identify your role. Parent, advocate, or attorney. District employees are not permitted to use this platform to surveil families — submitting a report is an attestation that you are one of the listed roles.
5 Write what happened. Stick to facts you observed. The narrative should be at least 30 characters. Aim for 30-500 words.
6 Acknowledge the defamation notice. Four clauses covering truth, opinion, good faith, and responsibility. Type your initials in each box.
7 Submit. The system runs PII scrubbing, then sends the scrubbed text to two AI models (Claude and Gemini) for defamation-risk review. You'll see one of three outcomes: approved (published to the feed), queued (held for human review), or rejected (not published).

How the privacy mechanics work

Two-stage scrubbing

First pass is deterministic — email addresses, phone numbers, URLs, and titled names (e.g. "Mrs. Smith") are replaced with placeholder tags. The district name you supplied is also explicitly removed if it appears in your narrative.

Second pass is AI — both Claude and Gemini independently scrub any remaining identifying information (first names of teachers, school nicknames, specific dates) and assign a defamation-risk score.

Defamation gate

Reports flagged as medium or high defamation risk by either model are NOT auto-published — they go to a human admin review queue regardless of the moderation verdicts. Reports flagged as low or clean by both models can be auto-published.

What's stored where

The public feed contains only:

The admin record additionally contains:

Admin-only fields are never sent to other subscribers. They exist so the admin can verify submissions, respond to legal demands if compelled, and audit moderation decisions.

Rate limits and abuse protection

Removing a submission

Right-to-delete: if you want a submission removed, email hello@sentineliep.com with the submission ID (you receive it after submitting). Removals from the public feed are usually completed within 48 hours. The admin record may be retained longer if required for legal compliance.

What this is and isn't

Community Reports are personal accounts, not verified facts. They reflect what individual parents and advocates experienced — they are not investigated, audited, or confirmed by SENTINEL·IEP. Read them as you would read any first-person testimony: useful context, not certified truth.
District employees are not permitted to use Community Reports to identify or surveil families. The role attestation at submission time, combined with terms of service, makes this an explicit prohibition. We will take action against any account found doing so.

Last updated: 2026-05-26