feat(metrics): enable repair tracking and Prometheus lifecycle for observability

feat(metrics): add repair_attempts_total and wire into repair loop outcomes

fix(core): non-zero generator timing + one-shot EMPTY retry; post-verify drop LIMIT to recover EM when ExecAcc=1

refactor(core): trace schema upgrade, verifier/executor sync, benchmark plot polish

feat(trace): enrich StageTrace (sql_length/row_count/verified/error_type/repair_attempts/skipped) and propagate in normalization; tag EmptySQL; annotate repair attempts

feat(core): refine pipeline & verifier; improve Spider benchmark accuracy

feat(core): always emit per-stage traces in Pipeline.run (fallback when StageResult.trace is empty)

feat(metrics): instrument per-stage and pipeline_total latency; count pipeline_runs_total (ok/error/ambiguous)

feat(trace): standardize StageTrace (add summary) and coerce duration_ms to int at API boundary

refactor(core): DI-ready Pipeline; add registry + YAML factory + typed trace/result

fix(pipeline): allow optional schema_preview in run() for better typing compatibility with benchmarks

fix(pipeline): align backend-frontend schema and stabilize SQL flow

build(mypy): fix type errors and add safety guards for None values

style: fix Ruff issues (F401 unused import, E741 ambiguous var)

style: format code with ruff

refactor: unify pipeline output via FinalResult model

Fix some typo

style: format code with ruff

init: NL2SQL Copilot base with API and Dockerfile