[INFRA 5] Adopt Alembic for database schema migrations

[m-khan-97]

Objective

Replace hand-rolled DDL with a real migration framework so schema changes are versioned, ordered, and reversible.

Current state

• Migrations are CREATE TABLE IF NOT EXISTS / ALTER TABLE statements embedded in api/models/finding.py (run_migrations), executed on every boot by startup.sh and create_app()
• No migration history table, no ordering guarantees, no rollback, no way to review a schema diff in a PR

Scope

• Introduce Alembic with a baseline migration capturing the current production schema
• Move all DDL out of DatabaseManager; it keeps only queries
• alembic upgrade head runs as the release/startup step (startup.sh and compose)
• Contributor docs: how to add a migration alongside a model change
• Verify baseline is a no-op against the existing Render database (stamp, don't recreate)

Out of scope

• Connection pooling → [FIX 2] Reliability: score endpoint 500, DB pooling, CVE timeout, query safety, thread safety #150 (REL-003)
• ORM adoption — raw SQL queries stay as they are

Acceptance criteria

• Fresh database reaches current schema via Alembic only
• Existing production DB is stamped and upgrades cleanly with zero data change
• run_migrations ad-hoc DDL removed
• CI passes, base branch dev

Category

Infrastructure / Database

[parthrohit22]

I'd Like to work on this.
PR will be ready by 7th July