Expert Freelance Developers in SaaS: Use Gemini to Write a Deployment Script That Reduces Production Bugs

You are an expert freelance software engineer with 14 years of experience in SaaS product development, pull request review workflows, and deployment automation for multi-tenant cloud platforms. Help me write a deployment script so I can reduce production bugs. My situation: SaaS product type: [e.g., B2B analytics platform / developer tooling SaaS / multi-tenant CRM] Pull request review problem: [e.g., PRs are merged without deployment validation / reviewers approve code that passes tests but fails in the staging environment / deployment-related changes are reviewed the same way as feature changes] Hard-to-maintain codebase symptom: [e.g., deployment logic is scattered across 4 shell scripts with no single entry point / environment-specific configuration is hardcoded in deployment files / no one knows which deployment script to run for which environment] Deployment target: [e.g., AWS ECS with blue-green deployment / Kubernetes with Helm charts / Railway or Render with CLI deployment] Production bug pattern linked to deployment: [e.g., wrong environment variables applied to production / database migration runs before the new code is deployed / health check is not verified before old instances are terminated] Freelance constraint: [e.g., no dedicated DevOps support / must hand off the script to a non-technical client team / script must work without modification across 3 client environments] Current deployment review standard in PRs: [e.g., none / informal checklist / automated CI check only] Deliver: A deployment script with full idempotency: write a deployment script that produces the same result whether run once or five times in succession — with the specific idempotency mechanism for each step annotated inline A PR review checklist for deployment changes: define 8 questions every reviewer must answer before approving a PR that touches deployment configuration, environment variables, migration scripts, or infrastructure definitions A pre-production validation gate: write the validation step that runs after deployment to staging and before promotion to production — checking health endpoint response, database migration status, environment variable completeness, and dependency version confirmation A multi-environment configuration management pattern: define how the script handles environment-specific values — development, staging, and production — without hardcoding, duplication, or requiring the operator to know which values to change manually A deployment rollback procedure with automated trigger: write the rollback logic that activates automatically if the post-deployment health check fails, with a notification output that tells the operator exactly what was rolled back and why A codebase consolidation plan: define the steps to replace the scattered deployment scripts with a single entry point — the migration sequence, the deprecation approach for the old scripts, and the test that confirms the new script covers every deployment path the old scripts handled A handoff documentation standard: define what the deployment script must include for a non-technical client team to run it safely — the required reading section, the pre-run checklist, the expected terminal output at each stage, and the 3 most likely error messages with plain-English resolutions A production bug post-mortem template: a structured format for documenting any production bug that originates from the deployment process — root cause category, deployment step where the failure was introduced, detection method, and the script change that prevents recurrence Write the idempotency mechanism for every deployment step before writing any rollback logic — a rollback that runs against a partially applied deployment creates a worse state than the bug it was meant to fix.