Connect Exchange scores to task generation:
1. Higher-scoring hypotheses should get more follow-up analyses, more tool invocations, more validation
2. Implement a scoring function: allocation_weight = composite_score * evidence_freshness * gap_coverage
3. When generating new tasks, weight task priority by the allocation_weight of the target hypothesis/entity
4. Add allocation_weight column to hypotheses table
5. Update the task creation logic in agent.py to use weighted priorities
Completion Notes
Auto-release: non-recurring task produced no commits this iteration; requeuing for next cycle