Nonfiction Generation Pipeline
==============================
How the workflow CHANGES based on Purpose × Structure
INPUT PHASE
-----------
User provides: concept, purpose (why read), category (subject), optional preferred framework
Example inputs:
- "Leadership for introverts" + PURPOSE=TRANSFORM + CATEGORY=LEADERSHIP
- "How to code in Python" + PURPOSE=LEARN_HANDS_ON + CATEGORY=TECHNOLOGY
- "Why nations fail" + PURPOSE=DECIDE + CATEGORY=HISTORY
CLASSIFICATION PHASE (NEW)
---------------------------
System maps: purpose → framework families → specific framework
if PURPOSE == "learn_hands_on":
framework = select("tutorial" or "howto")
stages = ["prerequisites", "step 1", "step 2", ...]
elif PURPOSE == "transform":
framework = select("transformation_journey" or "mountain_structure")
stages = ["wake-up", "denial", "dark night", ...]
elif PURPOSE == "decide":
framework = select("big_idea" or "problem_solution")
stages = ["problem", "evidence", "solution", ...]
SCAFFOLDING PHASE (CHANGES)
---------------------------
Different STAGES based on framework
TUTORIAL scaffold:
- Introduction → Prerequisites → Step 1 → Step 2 → Step 3 → Completion → Next Steps
TRANSFORMATION_JOURNEY scaffold:
- Wake-Up → Denial → Dark Night → Realization → Path → Struggles → Breakthrough → New Normal
BIG_IDEA scaffold:
- Promise → Opposition → Evidence → Implications → Counter-Arguments → Conclusion
DRAFTING PHASE (CHANGES)
------------------------
Different AGENTS activated based on purpose
if PURPOSE == "learn_hands_on":
agent = "TutorialWriter" # Focus on clarity, exercises, checkpoints
tone = "encouraging, clear, patient"
elif PURPOSE == "transform":
agent = "TransformationWriter" # Focus on emotion, narrative, inspiration
tone = "empathetic, honest, motivational"
elif PURPOSE == "decide":
agent = "EvidenceWriter" # Focus on data, proof, credibility
tone = "authoritative, data-driven, persuasive"
CRITIQUE PHASE (CHANGES)
------------------------
Different EVALUATION CRITERIA based on purpose
if PURPOSE == "learn_hands_on":
check: "Can a reader actually complete the steps?"
check: "Are prerequisites clear?"
check: "Is there a sense of progression?"
elif PURPOSE == "transform":
check: "Does it feel emotionally honest?"
check: "Is the transformation arc believable?"
check: "Would this inspire change?"
elif PURPOSE == "decide":
check: "Is the evidence credible?"
check: "Are counter-arguments addressed?"
check: "Does it lead to a clear recommendation?"
OUTPUT PHASE (CHANGES)
----------------------
Different FORMAT based on purpose
if PURPOSE == "learn_hands_on":
format = "markdown_with_code_blocks, exercises, checkpoints"
elif PURPOSE == "transform":
format = "narrative_chapters, emotional_arcs"
elif PURPOSE == "reference":
format = "indexed_sections, cross_references, searchable"