[Exchange] Extend market pricing to all artifact types done

Goal

Generalize the LMSR market pricing model from hypotheses-only to all artifact types.
Every artifact in the registry should have a market_price reflecting collective quality
assessment, with price history tracking all movements and their causes.

Current State

• market_dynamics.py (1275 lines): LMSR-inspired pricing for hypotheses only
• hypotheses.market_price column exists (0-1, default 0.5)
• price_history table: 7,200 entries tracking hypothesis price movements
• market_transactions table: 3,765 entries (all score_update actions)
• artifacts table has quality_score but no market_price

Acceptance Criteria

- experiment: extraction_confidence replication_status citation_count
- kg_edge: evidence_strength source_quality confirmation_count
- paper: citation_count journal_impact recency
- model: prediction_accuracy validation_score usage_count
- dataset: completeness provenance_quality usage_count
- wiki: edit_count reference_quality link_density
- figure/notebook/code: usage_count parent_quality freshness

Approach

Migration: Add market_price to artifacts table, seed from quality_score

Extend price_history to support artifact_id (not just hypothesis_id)

Build type-specific price factor functions

Create update_artifact_prices() batch function for periodic updates

Wire into artifact registration (initial pricing) and quality propagation (price updates)

Add API endpoints and UI integration

Dependencies

• None (foundation task — can start immediately)

Dependents

• exch-qm-02-PART — Participant agents need prices to evaluate
• exch-qm-03-LIFE — Lifecycle transitions trigger on price thresholds
• exch-qm-06-FEED — Debate feedback needs price adjustment mechanism

Work Log