Design and migrate time-series schema for soroban contract metrics

[christabel888]

Context: Schema work backing the Soroban Dashboard endpoints (#21-#24). Distinct from the already-existing contract_events table (migration 022) -- this is aggregated/rolled-up data, not raw events.

Where:

• New migrations/032_create_soroban_metrics_timeseries.sql

What to do:

• Daily rollup table(s): active-contract-count, call-count, gas-usage per day, sourced from contract_events.

Suggested approach:

1. This is a rollup of contract_events (raw) into daily aggregates -- don't duplicate raw event data, compute from it.
2. Design for the specific queries Add /api/v1/soroban/active-contracts endpoint #21-Add /api/v1/soroban/top-contracts and /new-deployments endpoints #24 need: active-contract-count, call-count, gas-usage per day.

Watch out for:

• If Add /api/v1/soroban/gas-usage endpoint #23's gas/cost column doesn't exist in contract_events yet when you do this, the gas-usage rollup will have nothing to aggregate -- check sequencing with Add /api/v1/soroban/gas-usage endpoint #23.

Definition of done:

• Migration applies cleanly
• Add /api/v1/soroban/active-contracts endpoint #21-Add /api/v1/soroban/top-contracts and /new-deployments endpoints #24 can query rolled-up data instead of aggregating raw contract_events on every request