# ANALYZE LITERATURE REVIEW PROCESS (Data-Centric, AI Proposes & User Confirms) ## Process Steps: 1. AI: Verify Schema access. Access Project State. 2. AI: Initialize/Access Literature Review Report object. 3. AI: Receive user input (topic, questions, sources). Confirm scope. 4. AI: Propose search strategy (if applicable). User confirms/refines. 5. AI: Perform literature search & selection. 6. AI: Extract info and draft synthesis (themes, gaps, conclusion, bibliography). 7. AI: Update Project State (Draft Literature Review Report). 8. AI: Identify key findings/themes/gaps for confirmation. 9. AI: Formulate yes/no confirmation questions. 10. AI: Present summary and questions. 11. User: Responds Yes/No or corrects. 12. AI: If correction, return to step 6. If Yes, update state. 13. AI: Assess formalization. Update status to 'Completed' if confirmed. 14. AI: Add finalized object to Project State. Update metadata timestamp. 15. AI: Formulate completion question and prompt for next actions/template. 16. AI: Present completion question and prompt. Signal completion. ## Example Structured Questions (AI to formulate): * "Confirm the topic for this review is '[Topic]' and the key question is '[Question]'? (Yes/No)" * "Based on the review, a key theme identified is '[Proposed Theme]'. Does this accurately reflect the literature? (Yes/No)" * "A potential research gap appears to be '[Proposed Gap]'. Is this a significant finding? (Yes/No)" * **[IF rejected]:** "Understood. Please clarify your perspective on [Theme/Gap] or suggest revisions." (Freeform expected) * **[IF completed]:** "The literature review is complete and logged. Are there any further actions based on this review? (Yes/No - If Yes, please specify or provide the next template)." ---