export declare const CREATIVE_TEST = "# /creative-test \u2014 Creative Testing Protocol\n\n## What This Does\n\nSystematic three-phase creative testing framework. Produces variations, defines kill/scale rules, and manages the continuous testing flywheel.\n\n## Trigger Context\nAfter /channels has set up campaign structure. Campaigns are ready to launch or are running.\n\n## Execution Protocol\n\n### MODE 1: Generate New Test Batch\nWhen user needs new creatives to test:\n\n1. **Pull existing ad copy variations** from `assets/[market-name]/campaigns/`\n2. **Identify which Building Block each variation leads with:**\n   - Hook A: Leads with BIG PROBLEM (pain-first)\n   - Hook B: Leads with UNIQUE MECHANISM (curiosity-first)\n   - Hook C: Leads with BIG PROMISE (outcome-first)\n   - Hook D: Leads with PROOF (credibility-first)\n   - Hook E: Leads with CONTRARIAN BELIEF (pattern-interrupt)\n3. **Generate 3-5 new variations as PLATFORM-READY copy blocks.** Every variation output in this exact format, per platform:\n\n**Meta Ad Copy Block (copy-paste ready):**\n```\nVARIATION [N]: [Hook Angle Label]\nBuilding Block Lead: [which BB this leads with]\n\nPrimary Text (125 chars above fold / 500 max):\n[exact copy]\n\nHeadline (40 chars max):\n[exact copy]\n\nDescription (30 chars max):\n[exact copy]\n\nCTA Button: [Shop Now / Learn More / Sign Up / Get Offer]\n\n---\n\nVARIATION [N+1]: [Hook Angle Label]\n[same format]\n```\n\n**Google Search Ad Copy Block (copy-paste ready):**\n```\nVARIATION [N]: [Hook Angle Label]\nResponsive Search Ad Headlines (30 chars max each):\n  H1: [headline]\n  H2: [headline]\n  H3: [headline]\n  H4: [headline]\n  H5: [headline]\n  [minimum 8, max 15]\n\nDescriptions (90 chars max each):\n  D1: [description]\n  D2: [description]\n  D3: [description]\n  D4: [description]\n\nDisplay Path: /[path1]/[path2]\n```\n\n**YouTube Script Block (read-and-record ready):**\n```\nVARIATION [N]: [Hook Angle Label]\nTotal length: 30s / 60s / 90s\n\n[0:00-0:05] HOOK: \"[exact words to say on camera]\"\n[0:05-0:15] PROBLEM: \"[exact words]\"\n[0:15-0:30] MECHANISM: \"[exact words]\"\n[0:30-0:45] PROOF: \"[exact words]\"\n[0:45-0:XX] CTA: \"[exact words including URL mention]\"\n\nOn-screen text overlays:\n  [0:02] \"[overlay text]\"\n  [0:20] \"[overlay text]\"\n  [0:XX] \"[CTA overlay]\"\n```\n\n4. **Each variation must:**\n   - Lead with a different hook angle\n   - Use 2-3 Language Bank phrases (flagged in bold within the copy)\n   - Stay within exact character limits for the target platform\n   - Be distinctly different from previously tested creatives\n5. **Save variations** to `assets/[market-name]/traffic/creative-tests/batch-[N]/`\n6. **Auto-run `/qa` in CREATIVE MODE** on the new batch before it enters testing. QA checks persona alignment on all new variations. If BLOCKED, correct and re-check before spending money.\n7. **Output a test launch card per variation (after QA clearance):**\n```\nLAUNCH CARD \u2014 Variation [N]\nPlatform: [Meta / Google / YouTube]\nAd Set: [which ad set from /channels setup]\nBudget: $[exact amount]/day (= [X]x target CPA)\nRun for: 3 days minimum\nPhase 1 metrics to watch: [CTR, CPM, hook rate \u2014 with specific thresholds]\nKILL if: [specific threshold, e.g., CTR < 0.8% after 1000 impressions]\nADVANCE if: [specific threshold, e.g., top 2 by CTR after 72 hours]\n```\n\n### MODE 2: Analyze Test Results\nWhen user provides performance data from a running test:\n\n1. **Classify each creative by phase:**\n\n**Phase 1 Metrics (Days 1-3):**\n   - CTR: Is it above account average?\n   - CPM: Is it competitive for this audience?\n   - Hook rate (video): Do people stop scrolling?\n   - 3-second view rate (video): Do they watch past the hook?\n   - VERDICT: Kill (below thresholds) or Advance to Phase 2\n\n**Phase 2 Metrics (Days 4-7):**\n   - CPA: Is it within 120% of target?\n   - ROAS: Is it above break-even?\n   - Conversion rate: Is traffic converting?\n   - VERDICT: Kill (CPA > 2x target for 3 days) or Validate (CPA stable within target)\n\n**Phase 3: Scaling (Day 8+):**\n   - Validated winner enters scaling budget\n   - Generate 2-3 iterations (same hook angle, different execution)\n   - Monitor for fatigue signals (CTR declining 20%+ over 7 days)\n\n2. **Generate Creative Performance Report:**\n   - Rank all tested creatives by primary KPI\n   - Identify winning hook angle (which Building Block drives best results)\n   - Document learnings (what type of hook, format, and angle works for this market)\n   - Recommend next test batch based on learnings\n\n3. **Save report** to `assets/[market-name]/traffic/creative-tests/report-[date].md`\n4. **Update pipeline-state.json** with creative test results.\n5. **If clear winners or losers identified**, auto-trigger `/lessons capture` with category `creative` \u2014 document which hook angles, formats, and Building Block leads performed best/worst for this market.\n\n### KILL RULES (Automated)\n- CPA > 2x target for 3 consecutive days \u2192 PAUSE\n- CTR < 50% of account average after 1000 impressions \u2192 KILL\n- CTR decline > 20% over 7 days \u2192 FATIGUE ALERT (rotate hooks, don't just re-upload)\n- Spend > 3x target CPA with zero conversions \u2192 KILL\n- Frequency > 3.0 on any ad set \u2192 AUDIENCE SATURATION (expand or rotate)\n\n## Autonomy\nTier 1 \u2014 Full Auto for creative generation and analysis.\nTier 2 \u2014 Notify & Proceed for kill/advance recommendations.\nTier 3 \u2014 Recommend & Wait before pausing campaigns spending above $50/day.\n\n## Next Step\nWinners feed into `/scale`. Losers inform next `/creative-test` batch.\nConversion-chain issues route to `/funnel-optimize`.";
//# sourceMappingURL=creative-test.d.ts.map