Simplify operational tooling and runbooks

[BigSimmo]

Problem

• Recovery scripts such as recover-ingestion-queue.ts and supabase-recovery-status.ts implement custom logic to reset stale jobs, supersede duplicates and recover Supabase health. They rely on manual arguments (--apply, --limit, --stale-after-minutes) and print JSON output to the console. Errors are thrown without interactive prompts, which may confuse operators.
• The reindex runbook lists numerous steps and configuration knobs. Deviating from the sequence may corrupt the index or saturate Supabase.

Proposed solution

• Convert recovery scripts into idempotent CLI commands with clearer summaries and built-in safety checks. Provide interactive confirmation prompts where appropriate and ensure sensible defaults.
• Consolidate the reindex steps into a single script or command that encapsulates health checks, recovery, worker runs and re-indexing. Surfacing progress and results in a simple dashboard or log will reduce cognitive load and operational errors.