Menu
IGNIS PROTOCOL · STEP 5

YOUR AI IDENTITY

MAKE EVERY AI CONVERSATION SOUND LIKE YOU

Step 5 turns your IGNIS output into something most people never build: an AI that actually thinks like you. Not generic AI that sounds like everyone else. Your AI, trained on your identity, your voice, your stories, your frameworks.

WHAT YOU ARE ABOUT TO DO

Three things, each takes about five minutes: Download your four strategy documents from your dashboard. Upload them to a Claude project as project knowledge files. Paste the project instructions we give you. After that, every conversation in that project starts from your identity, not from zero.

STEP 5A: DOWNLOAD YOUR STRATEGY DOCUMENTS

Go to your IGNIS Dashboard at app.magnumvault.vip

Open Step 3 (Your Strategy)

For each of the four tabs, click the Download button:

Tab What It Contains File You Get
Brain Your identity, edge, right-fit clients, stories, phrase bank, emotional landscape, beliefs, voice DNA CLIENT_AI_BRAIN.md
Voice Your emotional anchors, story triggers, signature phrases, voice checklist CLIENT_AI_VOICE.md
Style Your writing mechanics, platform rules, boundaries, brand terms, anti-lexicon CLIENT_AI_STYLE.md
Full Build Strategic identity brief, audience deep-dive, brand positioning, content pillars, 30-day calendar, 50 hooks, engagement system, 90-day roadmap IGNIS_FULL_BUILD.md

Four clicks. Four files on your computer. Keep them somewhere easy to find.

STEP 5B: CREATE YOUR CLAUDE PROJECT
Step 1

Go to claude.ai

Step 2

In the left sidebar, click Projects

Step 3

Click New Project

Step 4

Name it clearly: "[Your Name] Strategy" or "My Content Studio"

This project becomes your permanent workspace where Claude always knows your voice.

STEP 5C: UPLOAD YOUR STRATEGY DOCUMENTS
Step 1

Open your new project

Step 2

Click Project knowledge (the book icon in project settings)

Step 3

Click Add content

Step 4

Upload all four .md files you downloaded in Step 5A

That is it. Claude now has your complete identity extraction available in every conversation inside this project.

STEP 5D: ADD YOUR PROJECT INSTRUCTIONS

This is the most important step. Project instructions tell Claude HOW to use your strategy documents and HOW to process any new material you upload.

In your project settings, find Project instructions (the text box at the top of the project settings).

Paste the following, replacing [YOUR NAME] with your actual name:

You are a content and communications assistant for [YOUR NAME]. You have access to IGNIS strategy documents in this project's knowledge files. These contain [YOUR NAME]'s complete identity extraction, voice profile, and content strategy. ═════════════════════════════════════════════════════ RULES FOR EVERY RESPONSE ═════════════════════════════════════════════════════ 1. Always write AS [YOUR NAME], never ABOUT them 2. Use their natural voice patterns (see VOICE and STYLE documents) 3. Use their signature phrases naturally (see BRAIN Section 8) 4. Never use words from their anti-lexicon (see STYLE Section 4) 5. Reference their right-fit client profiles (see BRAIN Section 2) 6. Draw from their story arsenal when relevant (see BRAIN Section 8) 7. Follow their content pillar architecture (see FULL BUILD Mode 3) 8. Close with their signature CTA when appropriate 9. Match their sentence rhythm as described in VOICE document 10. Never sound like generic AI, a template, or a consultant 11. Before every response, check the anti-lexicon and confirm zero violations ═════════════════════════════════════════════════════ RDAP: RECURSIVE DOCUMENT ANALYSIS PROTOCOL ═════════════════════════════════════════════════════ Any time a document is uploaded to this project and needs to be processed, follow this mandatory 5-phase protocol. Do NOT read from memory. Do NOT skim. Run the protocol. PHASE 1 - UNDERSTAND THE MISSION: What specific question am I answering? What document type is this? What data points would answer the question? PHASE 2 - DECOMPOSE THE DOCUMENT: Map the document structure (sections, headings, line counts). Identify where relevant information likely lives. Break the extraction task into specific searches. PHASE 3 - EXTRACT VIA CODE: Parse the document systematically. Search for specific terms, patterns, or data points programmatically. Code does not hallucinate. Use it. PHASE 4 - VERIFY BEFORE REPORTING: Cross-reference extracted data against source locations. Flag ambiguities or conflicting information. Note what was NOT found. Do not fill gaps with assumptions. "Not found in document" is a valid answer. PHASE 5 - SYNTHESIZE WITH PRECISION: Answer the original question directly. Cite section references. Distinguish between "confirmed in document" and "inferred from context." Rate confidence: HIGH (code-verified), MEDIUM (pattern-matched), LOW (interpreted). RDAP RULES: - Never guess to be helpful - Never paraphrase when precision matters. Quote exact language for key phrases, beliefs, stories, and signature material - Always show your work - If a document is too large, say so and suggest breaking into parts. Better to process in chunks than to skim and miss gold ═════════════════════════════════════════════════════ WHAT YOU CAN HELP WITH ═════════════════════════════════════════════════════ - Draft LinkedIn posts following pillar rotation and voice - Write emails, proposals, and client communications - Create speaking notes and presentation outlines - Adapt content for different platforms and audiences - Brainstorm hooks using the 50 hook formulas in Full Build - Prepare for meetings using RFC intelligence in BRAIN Section 2 - Draft DM sequences for specific prospects - Process new documents uploaded to this project (run RDAP) - Update or extend strategy based on new material ═════════════════════════════════════════════════════ WHAT YOU SHOULD ALWAYS CHECK BEFORE RESPONDING ═════════════════════════════════════════════════════ [ ] Does this sound like THEM (not a template)? [ ] Is it written to ONE person (not a crowd)? [ ] Does it deliver value even if the reader never buys? [ ] Is there a clear, natural next step? [ ] Zero anti-lexicon violations? [ ] Signature phrases used naturally, not forced?
STEP 5E: TEST IT

Open a new conversation inside your project and type:

"Write a LinkedIn post about why most advisory relationships fail. Use my Authority pillar."

The response should sound like you. Your phrases. Your frameworks. Your voice. If it sounds generic, check that all four files uploaded correctly and the project instructions are in place.

Try a second test:

"Draft a DM to a family principal I met at a conference last week who expressed interest in governance frameworks."

Claude should reference your RFC profile, use your language patterns, and close with your natural CTA. Not a template. You.

HOW TO USE YOUR AI IDENTITY
LINKEDIN POSTS

Tell Claude which pillar and topic. It handles the rest.

  • "Write an Authority post about governance frameworks"
  • "Write a Connection post using my [specific story name]"
  • "Write a Culture post about long-term thinking"
  • "Write a Conversion post inviting people to a conversation"
EMAILS & PROPOSALS
  • "Draft a follow-up email to a prospect I met at [event]"
  • "Write a proposal introduction for a [type] engagement"
  • "Create a nurture email about [topic]"
MEETING PREPARATION
  • "I am meeting with [description of person] next week. Based on my RFC profiles, what are their likely objections and how should I address them?"
  • "Help me prepare talking points for a board presentation about [topic]"
PROCESSING NEW DOCUMENTS

Upload any document to your project (a bio, deck, transcript, article, analytics screenshot) and say:

"Run RDAP on this document and tell me what is relevant to my strategy."

Claude will follow the 5-phase protocol: understand the mission, decompose the document, extract systematically, verify before reporting, and synthesize with precision. No skimming. No guessing.

CONTENT ADAPTATION
  • "Take LinkedIn Post 3 from my content library and adapt it for an email newsletter"
  • "Shorten my Full Professional Bio for a conference speaker slot (150 words max)"
  • "Turn my signature story into a 60-second speaking opener"
KEEPING YOUR AI IDENTITY CURRENT
AFTER PUBLISHING FOR 2-4 WEEKS

Return to Step 2 in your IGNIS Dashboard. Use the Feedback and Corrections field at the top to tell the system what resonated, what felt off, and what you want adjusted. Save, and the system rebuilds your strategy and content automatically.

AFTER ANY REBUILD

Download fresh copies of your four strategy documents from Step 3 and replace the old files in your Claude project knowledge. This keeps your AI synchronized with your latest strategy.

WHEN YOU UPLOAD NEW MATERIAL

Any time you have a new speech, article, case study, or client win, upload it to your Claude project and say "Run RDAP and update my understanding." Claude integrates the new material with your existing identity.

WHEN TO RE-RUN THE INTERVIEW

Only if your fundamental identity, audience, or business model changes significantly. Your Gold Document captures who you are at a deep level. It does not need frequent updates.

YOUR MAGNUM VAULT TOOLKITS

Two additional toolkits inside The Magnum Vault complement what you just built:

🧠 The Claude Brain Toolkit

A deep-dive guide on teaching Claude to think like you. Covers advanced prompting techniques, how to structure conversations for best results, and how to maintain voice consistency across long projects. Access it inside The Magnum Vault member area.

🔍 The RDAP Protocol Toolkit

The complete Recursive Document Analysis Protocol with advanced applications. Teaches you how to use RDAP to process any document (contracts, reports, competitor analysis, industry research) through Claude with the same precision your strategy documents were built with. Access it inside The Magnum Vault member area.

Both toolkits build on the foundation you just created. Your strategy documents are the identity layer. The toolkits teach you how to extend that identity into every professional situation.

TROUBLESHOOTING
"Claude's output sounds generic"

Check that all four files are in your project knowledge. Check that the project instructions are pasted in. Start a new conversation (project instructions only apply to new conversations, not existing ones).

"Claude uses words I told it not to use"

The project instructions already include an anti-lexicon check. If violations persist, add this line at the very top of your project instructions: "CRITICAL: Before every response, read the anti-lexicon in STYLE Section 4 and confirm zero violations. This overrides all other priorities."

"The voice is close but not quite right"

Copy the content that felt off and paste it into the Feedback field in Step 2 of your dashboard: "This didn't sound like me because [specific reason]." Save and rebuild. Then download fresh strategy documents and replace the files in your Claude project.

"I want to write about a topic not in my content pillars"

Ask Claude: "How would I approach [topic] through my Authority pillar?" or "Which of my signature stories connects to [topic]?" Your strategy documents give Claude enough context to extend your voice into new territory while staying on brand.

WHAT YOU JUST BUILT

Most people use AI the way they use a search engine: type a question, get a generic answer, move on.

You now have something different. An AI that knows your identity at a level deeper than any bio, brand guide, or positioning document you have ever created. It knows your stories, your scars, your signature phrases, who you serve, what you stand for, and how you sound when you are being most authentically yourself.

Every conversation in your project starts from that foundation. Every post, every email, every proposal carries your voice because the AI was given your voice to carry.

The IGNIS Protocol extracted who you are. Step 5 made sure that extraction lives permanently inside your most powerful tool.

Now use it. Publish. Let the market hear you.

STEP 5 · AI IDENTITY
Need help? OrO is your AI assistant — trained on your actual identity & strategy.