feat: HTTP health/readiness endpoints for container deployments #1577

Problem

When Squad runs in Kubernetes or Azure Container Apps, operators need liveness and readiness probes for production-grade deployments. Currently:

• Squad does not expose any HTTP server
• There is no /healthz or /readyz endpoint
• There is no SQUAD_HEALTH_PORT environment variable

Kubernetes restarts exited containers automatically (via restart policy), which provides a basic safety net. However, platform-level readiness signaling — preventing traffic/jobs from being routed to a pod that hasn't finished startup and connected to the GitHub API — is not possible without an explicit endpoint.

Proposed Contract

SQUAD_HEALTH_PORT=<port>  # when set, start a lightweight HTTP server

Why This Matters

The container-image.md canonical docs page and both ACA/AKS scenario pages reference this as a planned feature. Without it, operators must rely on container restart policy alone or use exec-based process probes.

Prerequisite For

• docs PR docs: Azure hosting guides — Dockerfile contract, ACA, and AKS runbook #1575 (Azure hosting guides) — links this issue as a future improvement
• Any ops/telemetry runbook that depends on readiness signaling

Non-Goals

• Full metrics/observability server (separate concern — see OTEL integration)
• Status dashboard