feat(prisma): add query-driven indexes and remove redundant ones

[Kaylahray]

Summary

Enhance the Prisma schema with safer, query-driven indexes to improve data integrity and database performance without introducing breaking data changes.
Closes #49

What Changed

• Removed redundant indexes already covered by unique constraints
• Added an index on Session.expiresAt to support expired-session cleanup
• Added a compound index on Position(userId, protocolName, status) for active position lookups
• Added an index on Position.status to support active position scans
• Added a compound index on YieldSnapshot(positionId, snapshotAt) for position history and latest snapshot queries
• Added a compound index on ProtocolRate(protocolName, assetSymbol, fetchedAt) for latest rate lookups by protocol and asset

Why

These changes are based on the current Prisma query patterns in the codebase:

• session cleanup filters by expiresAt
• active positions are queried by status and by userId + protocolName + status
• yield snapshots are queried by positionId and ordered or filtered by snapshotAt
• protocol rates are queried by protocolName + assetSymbol and ordered by fetchedAt desc

Impact

• Improves query performance for the most common read paths
• Keeps existing data model behavior intact
• Avoids risky uniqueness changes that could break existing data
• Removes unnecessary duplicate indexes where unique constraints already provide index coverage

Notes

• Schema changes were completed in Prisma schema
• Migration generation/application was not completed in this pass due local Prisma/database environment issues
• No intentional breaking changes to existing data

[drips-wave]

@Kaylahray Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits