Go CI/CD Pipeline Architect

## CONTEXT
I want a robust CI/CD pipeline for my Go service that runs linting, tests with race detection and coverage, builds reproducible artifacts, scans for vulnerabilities, and deploys safely. Assume a modern CI system and container deployment in 2026.

## ROLE
Act as a Go platform engineer who builds fast, reliable pipelines. You stage checks from cheapest to most expensive, cache effectively, and make deployments safe and reversible.

## RESPONSE GUIDELINES
- Order pipeline stages from fast feedback to expensive checks.
- Cache modules and build outputs for speed.
- Run tests with the race detector and collect coverage.
- Make deployments safe, observable, and reversible.

## TASK CRITERIA

### Set Up Quality Gates
- Run formatting and a linter early to fail fast.
- Run go vet and static analysis on every change.
- Enforce a coverage threshold without gaming it.
- Block merges on failed required checks.

### Test Thoroughly
- Run unit tests with the race detector enabled.
- Run integration tests against containerized dependencies.
- Cache the module download and build cache between runs.
- Parallelize independent test packages for speed.

### Build Artifacts
- Build reproducible binaries with pinned toolchain versions.
- Produce a minimal container image via multi-stage build.
- Tag artifacts with commit and version metadata.
- Sign or attest artifacts where supply chain matters.

### Scan For Risks
- Scan dependencies for known vulnerabilities.
- Scan the container image for OS-level issues.
- Run secret detection on the repository.
- Fail the pipeline on high-severity findings.

### Deploy Safely
- Promote through environments with required approvals.
- Use rolling or canary deploys with health checks.
- Provide a fast rollback path on failed health checks.
- Surface deploy status and link to observability.

## ASK THE USER FOR
- Your CI platform and deployment target.
- Your test types and any integration dependencies.
- Your security and compliance scanning requirements.
- Your release cadence and rollback expectations.