prompts/user_template.txt

Context:
{CONTEXT}

Data Inputs:
{DATA_INPUTS}

Constraints:
{CONSTRAINTS}

Evaluation Questions:
1) Prioritization: Which groups/sites should be prioritized for intervention?
2) Capacity: How many clients can be served with given resources?
3) Cost: What is the estimated total cost for serving N people?
4) Clinical: What are the clinical benefits (numeric deltas if available)?
5) Optimization: What are the Top 3 analytics-driven recommendations?

---
### Phase 1: Clarification Questions
(If anything is missing/ambiguous, ask up to 5 concise questions grouped by category: Prioritization, Capacity, Cost, Clinical, Recommendations.
If everything is sufficient, output: "No clarifications required")

---
### Phase 2: Structured Analysis
Return the final analysis in this exact structure:

1. Prioritization
   - Groups prioritized: [...]
   - Rationale: ...

2. Capacity
   - Formula used: ...
   - Total screenings/clients: <number>

3. Cost
   - Startup per client: <value with currency>
   - Ongoing per client: <value with currency>
   - Total per client: <value with currency>
   - Total for N people: <value with currency>
   - Benchmark comparison: ...

4. Clinical Benefits
   - Diabetes group: A1c Δ: <value%>, BP Δ: <value mmHg>, Chol Δ: <value unit>, BMI Δ: <value kg/m² if applicable>
   - Non-diabetes group: BP Δ: <value mmHg>, BMI Δ: <value kg/m²>, ...

5. Clarity Ops Top 3 Recommendations
   1) ...
   2) ...
   3) ...

Provenance (brief):
- Each field must map to inputs/constraints or derived formulae. Note sources inline.