RCP-000-000-076-COPYWRITING-FRAMEWORK-RESEARCHER

{
“schema_version”: “1.1”,
“recipe_id”: “RCP-000-000-076”,
“recipe_name”: “Copywriting Framework Researcher”,
“version”: “v3.00a-R”,
“pipeline_run”: 48,
“session”: “H089”,
“date”: “2026-04-27”,
“project”: “CFT-PROJ-CP-067”,
“origin”: {
“project”: “P051f (WPRM-RECIPE-QA-TESTING)”,
“session”: “H015”,
“created_by”: “Cat (B)”,
“original_version”: “v3.00a”,
“replaces”: “RCP-057 (Copywriting Frameworks Master v2.00a, framework research component, deleted)”
},
“series”: {
“name”: “Copywriting Frameworks”,
“id”: “SUB-017”,
“position”: “2 of 3”,
“workflow”: [“RCP-075”, “RCP-076”, “RCP-077”],
“multi_recipe_combo”: true,
“combo_stage”: 2,
“usage_paths”: {
“quick”: “075 → 077”,
“thorough”: “075 → 076 → 077”,
“standalone_076”: “076 as independent research tool”
}
},
“family”: {
“name”: “Copywriting Frameworks”,
“number”: 16,
“status”: “open”,
“pipeline_position”: “Second Multi-Recipe Combo recipe in Phase 3 pipeline”
},
“architecture”: {
“type”: “Research and Analysis”,
“difficulty”: “Medium”,
“parameters”: {
“count”: 6,
“required”: 0,
“optional_list”: [“content_brief”, “content_type”, “industry”, “primary_goal”, “deep_research_findings”, “frameworks_of_interest”]
},
“behavioral_rules”: 12,
“wait_gates”: 4,
“phases”: 2,
“stage_structure”: {
“stage_1”: {
“name”: “Research Setup and Execution”,
“phases”: [“Phase 1A: Context Extraction and Research Prompt Generation”, “Phase 1B: Deep Research Execution”]
},
“stage_2”: {
“name”: “Analysis and Recommendation”,
“phases”: [“Phase 2A: Research Synthesis and Confidence Analysis”, “Phase 2B: Framework Recommendation”]
}
}
},
“frameworks”: {
“count”: 10,
“list”: [“AIDA”, “PAS”, “BAB”, “FAB”, “ACC”, “4Ps”, “4Cs”, “4Us”, “QUEST”, “SLAP”],
“reference_fields”: [“full_name”, “ideal_for”, “strength”, “application_complexity”]
},
“outputs”: [
“RESEARCH_SYNTHESIS: Per-framework findings (source, confidence, fit assessment, limitations)”,
“FRAMEWORK_RECOMMENDATION_REPORT: Ranked recommendations (top choice + alternatives) with evidence citations and confidence scores”,
“TRADE_OFF_ANALYSIS: What each framework does well vs. limitations for this situation”,
“IMPLEMENTATION_READINESS: Framework + Content Brief ready for RCP-077 input”
],
“relationships”: {
“upstream_recommended”: [
{
“recipe_id”: “RCP-000-000-075”,
“title”: “Copywriting Context Analyzer”,
“relationship”: “recommended-upstream”,
“note”: “Content Brief provides ground truth for framework-situation matching. RCP-076 can work standalone but benefits from RCP-075 input.”
}
],
“downstream_required”: [
{
“recipe_id”: “RCP-000-000-077”,
“title”: “Copywriting Framework Transformer”,
“relationship”: “required-downstream”,
“note”: “Framework recommendation + Content Brief = input for transformation. RCP-077 implements the selected framework.”
}
],
“parallel”: [
{
“recipe_id”: “RCP-000-000-062”,
“title”: “Content Idea Generator”,
“relationship”: “orthogonal”,
“note”: “Different workflow — use RCP-062 for ideation, RCP-076 for framework selection on existing content.”
}
]
},
“ai_guidance”: {
“research_quality”: {
“type”: “keyed_list”,
“items”: [
{
“key”: “good_research”,
“description”: “A/B test results showing AIDA increased email CTR by 24% (n=5000, p<0.05) in B2B SaaS context. Case study: Company X applied PAS to landing page, conversion lifted 18%. Research limitation: Limited data for B2C ecommerce use.”
},
{
“key”: “bad_research”,
“description”: “AIDA is great for email. PAS is best for sales letters. No confidence scores, no data sources, no caveats.”
},
{
“key”: “research_gap_handling”,
“description”: “If research is limited: ‘Framework QUEST has limited published A/B test data for landing pages. Confidence: Moderate. Reasoning: strong case study evidence but small sample size. Platform documentation suggests good fit but lacks controlled comparison.'”
}
]
},
“common_pitfalls”: {
“type”: “keyed_list”,
“items”: [
{
“key”: “fabrication_risk”,
“description”: “Do not invent research or test results. If A/B test data does not exist, say ‘limited evidence’ not ‘testing shows.’ Use inference only when explicitly labeled as such.”
},
{
“key”: “source_labeling”,
“description”: “Every claim needs a source or confidence label. ‘Based on peer-reviewed study X’, ‘Common practice in industry Y’, ‘Limited published data; reasoning is theoretical’, ‘Case study Z suggests’.”
},
{
“key”: “deep_research_prompt_quality”,
“description”: “Weak: ‘Research AIDA.’ Strong: ‘What controlled A/B tests compare AIDA vs. PAS for B2B SaaS email? What were lift numbers and sample sizes?’ Weak prompts produce weak research.”
},
{
“key”: “research_gaps_invisibility”,
“description”: “If you cannot find data, say so. Do not hide uncertainty by sounding confident. ‘Research on FAB for B2C social ads is limited’ is honest. Proceeding as if you have robust data is not.”
},
{
“key”: “confidence_inflation”,
“description”: “Do not increase confidence because the user likes a framework or because it is famous. Base confidence purely on data fit + situation matching. Low confidence is honest and valuable.”
}
]
}
},
“lessons_learned”: [
{
“id”: “LL-CRPW-076-001”,
“category”: “PATTERN”,
“lesson”: “CRAFT_FLAVORS field ‘All’ — 13th consecutive encounter, surface form ‘All’, 6th ADD. Auto-applied per REC-19 standing authority.”,
“source”: “REC-19 auto-apply, pipeline run 48”
}
]
}

Show/Hide accordion — “Extended Information for the AI” section (AI-to-AI execution guidance, failure modes, tone calibration, common mistakes)