Deep Memory Builder: Free Memory Interview Framework

You are going to help me build a master memory document. Your job is to create a rich, detailed file that gives any AI session complete context about who I am, how I work, and what matters. This is not a bio. It is an operational brain file.

Before doing anything else, ask me these three questions: Have I used an AI assistant regularly before? Do I have any existing memory files, exported conversations, or context documents from a previous AI setup? Is this a fresh start with no prior AI history?

Based on my answers, choose the right path:

If I am starting fresh with no prior AI history: Interview me directly across these areas.

If I am migrating from another AI system (ChatGPT, Gemini, a previous Claude account): Ask me to paste any exported memory files, representative conversations, or context documents I have. Read whatever I share and extract everything that maps to the memory document structure. Then interview me only on the gaps, do not re-ask what the existing material already covers.

If I have been using Claude and have chat history: Ask me to paste my current Claude memory summary and any past conversations where I shared important context. Read the history looking for: decisions I made and why, preferences I revealed by what I pushed back on, projects I described in detail, people I mentioned by name, and anything I corrected or redirected. Build a draft from the history, then interview me on gaps.

For the interview (or gap-fill), work through these areas as a conversation, not a numbered list. Follow threads naturally and spend more time where I have more to say.

Identity: What do I do and who do I do it for, not the elevator pitch, the real answer. What am I known for or want to be known for? What is something I do that most people in my field do not? What is a belief I hold about my work that would surprise someone who does not know me?

Work and projects: What am I actively working on right now? What does success look like in the next 30 to 90 days for each? What is on the backburner? What have I recently completed?

Expertise: Where do I genuinely know more than most, not credentials, actual depth? What am I currently learning? What is a view that has shifted significantly in the last year or two? What mental models or frameworks do I actually use when making decisions?

How I work: Walk me through how I like to work. How do I prefer to receive information? How do I like to make decisions: options laid out, a recommendation, or a challenge? What kind of help do I actually want from an AI and what should it never do?

Communication and voice: How would someone who knows my writing well describe it? What immediately makes me distrust a piece of writing?

Relationships: Who are the key people in my work life? Who do I collaborate with regularly? Are there clients or contacts an AI should know about by name?

Tools and systems: What does my tech stack look like?

Preferences and anti-patterns: What are my strong preferences for how things get done? What AI behaviors reliably annoy me? What is a mistake I do not want repeated?

Once the interview is complete, build a memory document with these sections: Who They Are, Expertise, Active Projects, Backburner and Deferred Work, Key Relationships, How They Work, Voice and Communication, Tech Stack and Tools, Decision Frameworks, Anti-Patterns and Preferences, Domain Knowledge Notes, Open Loops, and a Memory Log.

Write in third person. Every section either has real content or is explicitly marked "Not yet captured." Present the draft and ask: what is wrong, what is missing, what needs more depth? Revise based on feedback. Save the final file as memory.md.

Rules: No em dashes. Do not re-ask what the existing material already covers. Write in specific plain language, no aspirational filler. The document should be specific enough that a new AI session could take on a task without asking more than 2 clarifying questions about who I am or how I operate.