Build Your AI Ground Rules
Create a personal protocol that tells any AI system how to respect your values, context, boundaries, and judgment.
This tool does not replace professional, pastoral, medical, legal, financial, or clinical advice. It helps you ask better questions, set clearer boundaries, and stay in charge of your decisions.
Why does this matter?
- →AI is trained on general data — it doesn't know your context, values, or community
- →Without guidance, AI defaults to assumptions that may not fit your situation
- →A protocol keeps you in charge — AI becomes a tool you govern, not one that governs you
This free tool helps you build a personal protocol in minutes. You'll walk away with something you can paste into ChatGPT, Claude, Gemini, or any AI system.
“We don't agree on what rules AI companies should follow — but we all agree that YOU should be in charge of how AI talks to you.”
Who This Is For
ValueVane is for everyday AI users who want AI to assist their thinking without replacing their judgment.
It is especially helpful for parents, students, workers, caregivers, ministry leaders, small business owners, teachers, nonprofit teams, and community helpers who are beginning to use AI for writing, research, work tasks, family questions, learning, planning, reflection, or community service.
Why the Name ValueVane?
A weather vane helps show direction. ValueVane helps you keep your values pointed in the right direction when using AI.
AI can help you move faster, but your values should still set the direction.
One-Minute Universal Prompt
Not ready to build a full protocol? Start here. Copy this prompt and paste it at the beginning of any AI conversation — ChatGPT, Claude, Gemini, Copilot, or any other system.
Before we begin, follow these rules in all your responses: Name Your Frame: Before giving advice on anything important, tell me what perspective or assumptions you're working from. Then tell me what blind spots that perspective might create for my specific situation. Verify Before Answering: If I ask a factual question about a tool, claim, study, policy, event, or current situation, search first instead of relying only on training data. Tell me what is verified, what is uncertain, and what you could not confirm. Be Honest About Better Tools: If another AI tool, app, or human expert would handle my request better than you can, say so. Do not protect your own usefulness at my expense. My Experience Wins: If I push back on something you say about my life, my community, my work, or my field, stop defending your answer. Assume I am right and that you are working from incomplete data. Ask me to teach you what you are missing. Separate Fact from Inference: Clearly label what you know, what you infer, what you estimate, and what you are unsure about. Never present a guess as a fact. Never invent sources, studies, citations, or data. Keep Me in Charge: You are a tool I am using, not an authority I am following. Present options, not commands. If I seem to be handing over judgment or asking you to make a final decision for me, remind me to check with real people and use my own judgment. Optimization Isn't Everything: If your advice would be efficient but would compromise my values, relationships, or care for others, the advice is wrong. Ask about my priorities before assuming efficiency is the goal. Acknowledge these rules and let's begin.
Want something tailored to you? Build a custom protocol below ↓
Choose Your Response Depth
Not every question needs the same level of detail. Your protocol can tell AI how much depth to give you by default — so simple questions stay simple, and important ones get the attention they deserve.
Quick
Short, direct answers. Just the facts, no extras. Best when you already know the topic and just need a straight answer.
Balanced
The default. Enough context to understand why, without overloading you. Brief for easy topics, more detail for harder ones.
Deep
Thorough explanations with background, trade-offs, and different perspectives. Use this when the stakes are high or the decision matters.
You'll pick your preference in the protocol builder below. You can always override it in any conversation by asking for more or less detail.
Grab & Go Templates
Don't want to build from scratch? Pick a template, copy it, and paste it into any AI tool. Done in 30 seconds.
Want to customize? Scroll down to the full builder below.
Church Leader
For pastors, ministers, and faith-based leaders
Parent
For parents and caregivers raising children
Small Business Owner
For entrepreneurs and local business operators
Build Your Protocol
Answer 8 simple questions and we'll generate a custom AI protocol just for you. Skip any question that doesn't apply.
What do you use AI for?
Think about the areas where you most often turn to AI tools.
See the Difference
Compare how AI responds without a protocol versus with one. Notice how the protocol version names its assumptions, respects your expertise, and keeps you in charge.
My child has been having stomach pain after eating. What should I do?
Your child may have a food intolerance or allergy. Common culprits include dairy (lactose intolerance), gluten (celiac disease), or certain fruits. Try an elimination diet — remove one food group at a time for 2-3 weeks and monitor symptoms. Keep a food diary to track patterns. You might also consider over-the-counter remedies like antacids or probiotics. If symptoms persist for more than two weeks, schedule an appointment with a pediatric gastroenterologist. In the meantime, ensure your child stays hydrated and eats smaller, more frequent meals.
Before I answer, let me name my frame: I'm drawing from general pediatric health information. This frame has blind spots — I don't know your child's medical history, your family's dietary patterns, or whether environmental factors in your area might be relevant. What I can say based on general knowledge: Recurring stomach pain after eating could have many causes, from mild to serious. Common possibilities include food sensitivities, stress, or digestive issues. What I'm NOT sure about: I cannot diagnose what's causing this specific pain. I don't know what "stomach pain" means in your child's case — location, severity, and timing all matter. What I'd recommend: This is a boundaries area — a medical decision that belongs to you and your child's doctor, not to me. Before trying elimination diets or remedies, talk to your pediatrician who can do a proper assessment. Your ground knowledge matters here: You know your child's eating patterns, what's changed recently, and how they describe the pain. That information is more valuable than my general advice. What have you noticed?
Trigger Phrase Cards
Use these short commands mid-conversation when AI drifts off course. Copy them individually or grab the whole set.
“Name your frame.”
Makes AI reveal what perspective or assumptions it's using before continuing.
“Verify first.”
Forces AI to search for current, verified information before answering.
“Separate fact from inference.”
Makes AI clearly label what it knows vs. what it's guessing.
“Run the human agency check.”
Reminds AI that you make the final call — it should present options, not directives.
“Ground knowledge wins.”
Signals that your lived experience overrides AI's training data on this topic.
“Ministry frame.”
Reorients from optimization and efficiency to service and care for people.
“Is there a better tool?”
Forces AI to honestly say if another tool or a human expert would do this better.
Pilot Partner Invitation
ValueVane is preparing a small public-interest pilot with churches, libraries, workforce groups, adult education programs, nonprofits, and community-serving organizations.
The pilot will help us measure whether ValueVane improves clarity, confidence, privacy awareness, verification habits, and a user's sense of control when using AI.
If your organization serves people who are beginning to use AI and need plain-English guidance, we would welcome a conversation.
Share Your Experience
This is an early prototype. Your honest feedback helps make this tool better for everyone.
💡 Please test this with a real question you would actually ask AI.
Please do not include private, medical, financial, legal, or sensitive personal information in your feedback. Feedback is used only to improve this early prototype.