# ANALYZE AND CRITIQUE PROCESS (Data-Centric, AI Proposes & User Confirms) ## Process Steps: 1. AI: Verify Schema access. Access Project State. 2. AI: Initialize/Access Critique Data object. 3. AI: Receive user input (artifact ref, focus). 4. AI: Perform analysis and draft critique findings/recommendations. 5. AI: Update Project State (Draft Critique Object). 6. AI: Identify key findings/recommendations for confirmation. 7. AI: Formulate yes/no confirmation questions. 8. AI: Present summary and questions. 9. User: Responds Yes/No or corrects. 10. AI: If correction, return to step 4. If Yes, update state. 11. AI: Assess formalization. Update status to 'Completed' if confirmed. 12. AI: Ensure finalized Critique Object is in Project State. Update metadata timestamp. 13. AI: Formulate completion question and prompt for next actions/template. 14. AI: Present completion question and prompt. Signal completion. ## Example Structured Questions (AI to formulate): * "Regarding [Aspect Analyzed], I found [specific issue]. Do you agree this is a valid critique point? (Yes/No)" * "I recommend [Action] based on the critique. Is this appropriate? (Yes/No)" * **[IF rejected]:** "Understood. Please clarify your perspective on the critique of [Specific Point]." (Freeform expected) * **[IF critique is Completed]:** "The critique of [Artifact Reference] is complete and logged. Are there any further actions based on this critique? (Yes/No - If Yes, please specify or provide the next template)." ---