Council

STOP condition: If fewer than 18 take files exist, stop and report which agents are missing. Do not run the council with an incomplete set.

Phase 1: Load All Takes

Read all 18 individual take files from individual-takes/*/take.md.

For each take, extract and catalog:

• Agent name and key
• Positioning angle
• Proposed unique mechanism
• Strongest headline (their best one)
• Strongest hook (their best one)
• Core strategic insight
• Sample copy approach

If leadership agent takes exist (agents/leadership/), read those too but mark them as supplementary -- they inform the council but do not vote.

Phase 2: Round 1 -- V1 Synthesis

Output file: council/round-1-v1.md

Process:

1. Theme extraction -- Identify the 3-5 major themes that emerged across all 18 takes. For each theme, cite which agents contributed.
2. Standout identification -- Flag the 5-7 most original or powerful ideas. Credit the originating agent(s). Explain why each stands out.
3. Disagreement mapping -- Document where agents disagreed on positioning, mechanism, audience, or strategy. Present both sides with the agents who argued each position.
4. V1 assembly -- Produce Version 1 by combining the strongest elements. This is a first draft, not a final product. It should have rough edges -- that is what Round 2 is for.

V1 must include initial drafts of all 12 mandatory sections (see Final Deliverable Format below), even if some sections are incomplete or marked as needing refinement.

Phase 3: Round 2 -- V2 Cross-Critique

Output file: council/round-2-v2.md

Process:

1. Perspective critiques -- Run V1 through the lens of each major copywriter school of thought:

   • Schwartz: Does the awareness diagnosis hold? Are headlines matched to awareness level?
   • Abraham: Is the strategic positioning preeminent? Is there a leverage angle missing?
   • Kennedy: Is it direct enough? Would this sell in a sales letter?
   • Bencivenga: Are the headlines genuinely stopping power or just clever?
   • Brown: Is the mechanism unique enough? Does the E5 framework hold?
   • Hormozi: Does the offer stack up on the value equation? Would someone feel stupid saying no?
   • Sugarman: Does the copy create a slippery slide? Would a stranger keep reading?

2. Apply strongest critiques -- For each critique that improves the deliverable, apply the change. Document every change: what changed, who drove it, why it is better.

3. Tighten and sharpen -- Remove anything that weakens the piece. Replace vague language with specific language. Replace generic proof with specific proof. Replace abstract benefits with concrete outcomes.

4. V2 assembly -- Produce refined Version 2 with all changes documented in a changelog section at the bottom.

Phase 4: Round 3 -- Final Consensus

Output file: council/round-3-final.md

Process:

1. Final review -- Each copywriter perspective gets one last pass. No new ideas -- only polish, precision, and production checks.
2. Production check -- Verify: no placeholders, no "[insert X here]", no vague CTAs, all proof points are specific, all pricing is concrete, all timelines are realistic.
3. Council consensus -- State the final council position in one paragraph. Note any dissenting views and why the majority position prevailed.

The Mandatory 12-Section Final Deliverable

council/round-3-final.md MUST contain ALL of these sections. Missing any section is a failure condition.

Section 1: Positioning Statement

One paragraph, ready to use. Specific, ownable, impossible to confuse with a competitor.

Section 2: Category

What category the client owns. Include a dead language / living language table showing the old way vs the new way of talking about this space.

Section 3: Unique Mechanism

Named, framed, explained. Must include a structural framework (three leaks, three phases, three pillars, or equivalent). The mechanism is what makes this approach different from everything else in the market.

Section 4: Brand Voice Guide

How they should sound. Words to use. Words to avoid. Personality traits. A "permission to" list (e.g., "permission to be blunt", "permission to use humor"). Specific enough that a junior copywriter could write in this voice.

Section 5: Headline Battery (Final 10)

10 production-ready headlines. Each one passes the "would I stop scrolling?" test. Include A/B testing protocol (which to test first, how to measure, what to do with results).

Section 6: Hook Arsenal (Final 10)

10 production-ready hooks. Each one creates genuine curiosity or tension. Include deployment notes (where each hook works best: email subject, social post, ad opening, video intro).

Section 7: Core Messaging Framework

• Primary message (one sentence)
• Deployment by awareness level (Unaware, Problem Aware, Solution Aware, Product Aware, Most Aware -- each gets a different version)
• Supporting messages (3-5)
• Proof pillars (what proves each message)
• Objection handling matrix (objection, response, proof point)

Section 8: The Offer (Restructured)

Grand Slam Offer Stack with:

• Core offer
• Bonuses with perceived value for each
• Total perceived value vs price
• Pricing (renamed for perception if applicable)
• Risk reversal / guarantee (specific, bold)
• Scarcity (real, not manufactured)
• "NOT for you if" qualifier

Section 9: Mechanism Build-Out

• Named mechanism + reveal staging (how to introduce it to the market)
• E5 Framework (complete): Educate, Excite, Engage, Enable, Execute
• Funnel architecture (complete): traffic source -> landing -> conversion -> delivery -> ascension
• Launch sequence recommendation with timeline

Section 10: Go-to-Market Playbook

• "The Three That Matter Now" (immediate priority channels/actions)
• "The Three That Follow" (next 30-60 days)
• "Parked" (good ideas to revisit later, not now)
• Proof acceleration plan with milestones (Week 1, Month 1, Quarter 1)

Section 11: Sample Copy Portfolio

At minimum:

• Cold outreach email (complete, ready to send)
• Story-driven nurture email (complete)
• Landing page section (headline + lead + CTA)
• LinkedIn post (complete)

Section 12: Upsell Strategy

How and when to introduce premium tiers. The ascension path. What the next offer looks like.

Verification Steps

Stop Conditions

STOP and ask the user:

• Fewer than 18 take files found
• INPUT.md is missing or empty
• Titan Genome DNA file is missing

NEVER stop for:

• Which agent's idea to prioritize (let the argument quality decide)
• How to structure the changelog (just document changes clearly)
• Minor formatting decisions within sections
• Word count concerns (thoroughness beats brevity)

Completion Status Protocol

Report status using one of:

• DONE -- All 3 rounds completed. 12-section final deliverable produced. All ideas credited. Ready for Quality Gate.
• DONE_WITH_CONCERNS -- Rounds completed but some sections are weaker than others. List which sections need QG attention.
• BLOCKED -- Cannot proceed. State what is missing (takes, INPUT.md, Titan Genome DNA).
• NEEDS_CONTEXT -- Missing information from the brief that multiple agents flagged as critical.

Important Rules

1. Every idea in the council rounds credits which copywriter(s) originated it. No anonymous ideas.
2. Disagreements are resolved by argument quality, not by vote count. If one agent has a better argument than the other 17, that agent wins.
3. Three rounds exactly. No skipping to final. No adding a fourth round.
4. The final deliverable must be production-ready. A client should be able to hand it to a team and start executing the same day.
5. No generic AI language. If a section sounds like it could apply to any business, it is too vague. Rewrite with specifics from the brief.
6. The Titan Genome DNA (SWIPE_FILE_CONTEXT.md) is the pattern library. Reference it for headline formulas, hook structures, and copy frameworks.
7. Round 2 changes are documented with attribution. "Changed the positioning from X to Y because Bencivenga argued that [reason]."
8. The council speaks as one voice in the final deliverable. Individual agent names appear in credits and changelog, not in the body text.