What problem does it solve? Deploying BigQuery agentic applications to Cloud Run without verification risks shipping broken conversational agents, missing table schemas, and unverified queries to production. This Skill enforces a mandatory 4-tier quality gate that blocks commits and deployments until all checks pass. ## Core Features & Use Cases - Tiered Verification Pipeline: Runs TDD unit tests for property graph DDL and A2A agent protocols, then an isolated local server E2E pipeline covering research, Medallion data generation, BigQuery deployment, verified query registration, and Gemini Enterprise A2A registration. - Git Safety & Automated Commit: Validates the working tree and automatically stages and commits changes when a commit message is provided. - Cloud Run Deployment with Smoke Checks: Builds and deploys the container to Cloud Run (bq-agents-v2 in us-central1) and runs a live smoke test against the deployed revision. - Use Case: Before merging a feature branch that adds a new reverse ETL continuous query, run the gate with a commit message to validate all tiers and deploy safely in one command. ## Quick Start Run the deployment gate with a commit message, for example: python3 scripts/deploy_with_e2e_gate.py --commit-msg "feat: add new verified query", or use --skip-deploy to validate without deploying.
