Safe HarborA calmer path through difficult momentsCONCEPT PROTOTYPE
Website access optionsChange the entire page. No diagnosis, account, or explanation required.
Public concept prototype For review and feedback
Safety without surveillance.
Help with harmful situations—in person and online

For young people facing harm—in person or online.

Safe Harbor is being designed for children, teenagers, and young adults who need a safer way to understand and explain a harmful situation. It would help organize what happened, protect unnecessary private details, compare possible next steps, and prepare to ask someone they decide may be able to help.

WHO IT IS FOR

Children, teenagers, and young adults—including neurodivergent and disabled young people—and the people supporting them.

WHERE HARM MAY HAPPEN

At school, at home, during activities, in public spaces, or through messages, social platforms, games, and online communities.

WHY IT EXISTS

To help someone create a clear account, understand choices, ask for support, and track follow-up without secret surveillance.

What “choosing someone” means

The young person selects a possible helper—such as a parent, caregiver, teacher, counselor, advocate, or safety professional—and previews exactly what that person would see. Nothing is sent automatically. The product cannot determine whether a particular person is trustworthy or safe.

Try the interactive demo How would it help me?
No secret monitoring No harmful media required Nothing shared automatically
Private by defaultOnly you can see this
You chooseNothing sent yet
Safe HarborFictional demo
AFTER SOMETHING HARMFUL HAPPENS

Know what to save and who can help.

Write it down, review it, then choose your next step.

1Record the contextDate, place or online space, notes✓
2Review and removeKeep only what feels safe→
3Choose who can helpNothing is shared automatically○
Not shared
Concept only. This is not monitored and cannot provide live emergency response. Immediate danger: call 911 or local emergency services.
What it would do

Four kinds of help during and after in-person or online harm.

It would not solve the case for you. It would help you slow down, organize the situation, and prepare for the next conversation.

1

Write it down

Note the date, place or online space, what happened, and what help is wanted.

You do not need to copy harmful pictures, messages, or exact language.
2

Check your notes

Read everything, fix mistakes, and remove details that feel too private or unsafe.

Every field would be optional.
3

Choose who can help

Compare choices such as a trusted adult, school, platform, counselor, or official help.

Nothing would be sent automatically.
4

Track what happens

Remember who you contacted, what they said, and whether you still need help.

You would keep one clear timeline.
In one sentence:It helps a young person make sense of an in-person or online incident, prepare to communicate, and keep track of follow-up. It does not watch, investigate, decide who is guilty, or contact anyone for them.
A fictional example

See what the app would actually help someone do.

No real person or incident.

This example follows a student who receives cruel group-chat messages. Nothing entered here is saved or sent.

STEP 01 OF 05 NOT SHARED

A student receives repeated cruel messages in a class group chat.

They open Safe Harbor and describe the situation in their own words. No names, screenshots, or account handles are needed for this example.

INFORMATION VISIBLE IN THIS STEP✓Approximate date✓Type of space✓Short description

Control ruleThe young person can stop or erase the draft at any time.

How it works

Four clear things the app helps you do.

It helps organize information and explain options. It does not investigate the incident, decide who is right, or contact anyone for you.

STEP 01

Make a simple, private record of the situation.

Add the date, place or online space, a link if one exists, and a short description. You do not need to copy harmful media or repeat cruel language.

✓Use your own words✓Every field is optional✓Save only what may help
BoundaryThe product could organize context and choices. It could not determine identity, guilt, intent, or whether a situation is safe.
What it can help with

In-person and online situations that may be hard to explain alone.

The app would help organize a young person’s own account of what happened in person or online. It would not prove an allegation, determine whether someone is currently safe, or replace emergency services or a qualified investigation.

01

Bullying or harassment

Cruel behavior, exclusion, humiliation, unwanted contact, or targeted conduct in person or online.

02

Threats or intimidation

Words, actions, messages, or posts that make someone fear for their safety.

03

Sexual harm, exploitation, or grooming

Pressure, manipulation, unwanted sexual behavior or contact, or requests for images.

04

Physical incidents or coercion

Unwanted physical behavior, coercion, restraint, property interference, or a harmful encounter.

05

Image sharing without permission

A record of where an image appeared and what the person knows about its spread.

06

School, organization, or platform follow-up

What was reported, when it was reported, and what response was received.

Choose who you are

What would this do for me?

Choose the description that is closest to you. Each answer explains the practical help in everyday language.

FOR CHILD

It helps you explain what happened without having to remember everything at once.

You could write a few simple notes, leave out anything you do not want to share, and see which trusted people may be able to help.

“You can stop, change your notes, or decide not to share.
Important

The app is not available yet. This website cannot take a report, store incident details, or send help.

Neurodivergent and disability access

Access needs should change the experience—not exclude someone from it.

Safe Harbor is intended to support neurodivergent and disabled young people across cognitive, learning, communication, sensory, visual, hearing, and motor access needs.

No diagnosis, disability label, or explanation should be required to use accessibility preferences.

01

Plain and predictable

Short instructions, one decision at a time, consistent controls, visible progress, and examples that explain unfamiliar terms.

02

Adjustable sensory load

Reduced visual density, limited motion, calm color choices, no flashing content, and no unnecessary sound or urgency cues.

03

More than one way to understand

Readable text, meaningful headings, screen-reader structure, optional summaries, and future alternatives to reading or typing alone.

04

Motor and vision access

Large controls, visible keyboard focus, zoom-friendly layouts, strong contrast targets, and operation without precise gestures.

05

No timer and no forced pace

Pause, go back, review, remove information, or stop. Access preferences should remain available throughout the experience.

06

Supported without hidden control

A support person may help when invited, but accessibility must not become secret monitoring, automatic disclosure, or loss of the young person’s voice.

AVAILABLE IN THIS CONCEPT

Testable access preferences

The full website now includes larger-text, higher-contrast, reduced-motion, and simplified-layout options. The interactive sandbox also includes plain-language, reduced-visual-load, and larger-text controls.

STILL REQUIRED

Independent testing with disabled users

This prototype does not claim accessibility conformance. It still needs assistive-technology testing, disabled-user research, cognitive walkthroughs, and remediation against documented findings.

DESIGN REFERENCES

Standards and cognitive guidance

W3C WCAG 2.2 overview ↗W3C cognitive accessibility guidance ↗

Privacy and safety

The boundary is the feature.

The app is meant to help someone think clearly and prepare for a conversation. It is not a surveillance, investigation, or emergency-response tool.

“

Before anything is shared, the young person selects a specific possible helper, sees exactly what that person would receive, removes any details they do not want included, or decides not to share.

Designed to keep

Only the minimum useful context.

The intended design would keep a young person’s own account, selected context needed for help, and a visible record of sharing choices.

  • The young person’s own words
  • Purpose-limited context
  • Visible sharing choices
Design rule · Not yet independently verified
01

No secret monitoring

It does not watch messages, accounts, location, or device activity.

02

No automatic reports

It does not contact parents, schools, platforms, police, or other organizations.

03

No guilt or identity decisions

It does not decide what happened, identify an offender, or promise to trace someone.

04

No emergency response

It is not monitored and cannot provide immediate protection during a crisis.

Intended information path

Every step should remain visible and interruptible.

Design target · Not yet verified
  1. 1
    Private reflectionStart with the young person’s own words.
  2. 2
    Purpose checkExplain why each detail could help.
  3. 3
    Sharing previewShow the recipient and exact summary.
  4. 4
    Visible recordPreserve the choice, purpose, and limits.

Local-first design goal: keep raw context on the device whenever the task allows.

SAFETY PROMISEWHERE IT STANDSWHAT WE CAN SHOW NOWWHAT STILL NEEDS REVIEW

Review before sharing

Shown in prototype

The fictional example includes a review and remove step.

Testing with young people under qualified safeguarding oversight.

No secret monitoring

Required rule

The prototype contains no background-monitoring feature.

An independent technical audit of the finished application.

Save less harmful content

Shown in prototype

The example records dates, links, and notes without copying harmful media.

Technical testing to confirm harmful media cannot enter through another path.

No identity guessing

Required rule

The product does not claim to identify or trace an alleged offender.

Testing designed to expose any accidental identity or guilt claims.

Crisis-aware routing

Not yet verified

Official help links and a clear emergency warning are displayed.

Independent review by qualified child-safety and crisis professionals.

Interactive concept demo

Explore an in-person or online scenario using fictional information.

Choose a made-up situation, decide what a summary should contain, and explore a mock next step. This browser-only demonstration does not save, analyze, upload, or send anything.

Fictional scenarios only

Do not enter names, handles, links, allegations, medical details, or information about a real incident. No free-text incident field is provided.

STEP 01 OF 04FICTIONAL DEMO · NOT SHARED
Make this demo easier to useChoose any combination. No diagnosis or disability disclosure is required.
Start safely

Which fictional situation would you like to explore?

These examples contain no real people, accounts, or incidents.

AVAILABLE NOW

Interactive concept sandbox

Visitors can test a fictional, browser-only version of the proposed flow.

NOT AVAILABLE

Real incident records or reporting

The demo does not store evidence, contact anyone, monitor emergencies, or operate as the safety app.

STILL REQUIRED

Independent review and product testing

Child-safety, privacy, legal, accessibility, security, and effectiveness review remain open.

TB
CREATOR & PRODUCT ARCHITECT

Tom Budd

Founder/CEO & CTO, ResoVerse LLC

LinkedIn profile ↗
Why this exists

Child safety should not require invisible surveillance.

Safe Harbor is being designed around a simple conviction: young people need practical ways to understand harm, preserve useful context, and ask for help without losing ownership of their information or being promised certainty technology cannot provide.

The project welcomes rigorous criticism from young people, caregivers, educators, accessibility specialists, child-safety professionals, privacy experts, and researchers.