Fix CI build/deploy/smoke pipeline for the self-hosted qa runner (#3)

## Summary
- Fix a chain of QA CI issues: Docker build/health-check flakiness, NuGet vuln pins, then adds the post-deploy integration + Playwright smoke suite and works through everything needed to make it actually run on the self-hosted `qa` runner (musl/Alpine job container, no node/dotnet/curl preinstalled, docker-outside-of-docker networking).
- Adds a fixed dev/QA seed admin user + fixed Development environment API key so integration tests and e2e specs have a stable target.
- Pins Aspire's `AppHost.cs` ports/credentials to match the docker-compose local dev defaults.
- Adds `tests/api/MicCheck.Api.Tests.Integration` coverage for disabled flags, environment-document bootstrap, identity override precedence, auth login, and unauthorized access; adds a Playwright e2e suite under `src/admin/e2e` (login, nav, context selection, features CRUD/toggle).
- Adds a `smoke-qa` CI job that runs both suites against the just-deployed QA stack, working around: no curl/node/dotnet on the bare runner, musl vs glibc (Playwright browsers run via the official `mcr.microsoft.com/playwright` image instead), and the runner's job-container network isolation (reach the QA stack via the docker bridge gateway IP; `docker cp` instead of a bind mount to get files into the playwright container, since paths don't cross the docker-outside-of-docker boundary).

## Test plan
- [x] Unit tests pass (dotnet test tests/api/MicCheck.Api.Tests.Unit, 243 passed)
- [x] API integration suite passes against the real QA stack in CI
- [x] Playwright e2e suite passes against the real QA stack in CI (verified 3/3 locally against a real dev API + vite server for the flakiest spec)
- [x] Full CI pipeline (build → deploy-qa → smoke-qa) green end to end on the qa runner

Reviewed-on: #3

scripts/ci/smoke-qa.sh

@@ -0,0 +1,55 @@
+#!/usr/bin/env bash
+# Post-deploy validation for the QA environment: runs the API integration
+# suite (real HTTP + real Postgres, seeding/cleaning up its own data) and the
+# Playwright admin e2e suite against the just-deployed QA stack. Intended to
+# run immediately after deploy-qa.sh, on the same self-hosted qa runner. This
+# job runs in its own job container, a sibling of the QA stack's containers
+# on docker's default bridge network - not "localhost" from the host's point
+# of view - so it reaches published ports via the bridge gateway IP, same as
+# deploy-qa.sh binds the db port to.
+set -euo pipefail
+cd "$(dirname "${BASH_SOURCE[0]}")" && source ./lib.sh
+cd "$CI_ROOT"
+
+ensure_dotnet
+ensure_node
+
+QA_ADMIN_PORT="${QA_ADMIN_PORT:-3001}"
+CI_HOST="$(docker_bridge_gateway)"
+[[ -n "$CI_HOST" ]] || fail "could not determine docker bridge gateway IP to reach the QA stack"
+BASE_URL="http://${CI_HOST}:${QA_ADMIN_PORT}"
+
+log "Resolved docker host as $CI_HOST for reaching the QA stack's published ports"
+
+export MICCHECK_API_BASE_URL="$BASE_URL"
+export MICCHECK_DB_CONNECTION_STRING="${MICCHECK_DB_CONNECTION_STRING:-Host=$CI_HOST;Port=55432;Database=miccheck;Username=miccheck;Password=password}"
+
+log "Running API integration suite against $BASE_URL"
+dotnet test tests/api/MicCheck.Api.Tests.Integration/MicCheck.Api.Tests.Integration.csproj -c Release --logger trx
+
+log "Installing admin e2e dependencies"
+npm --prefix src/admin ci
+
+# Playwright's bundled Chromium is a glibc binary and `--with-deps` only knows
+# apt - neither works on this musl/Alpine runner. Run the e2e leg inside
+# Microsoft's official Playwright image instead (glibc, browsers preinstalled
+# at /ms-playwright), as a sibling container reachable at the same bridge
+# gateway IP used above. Pin the image tag to the exact resolved
+# @playwright/test version so the test runner and browser build match.
+#
+# This script itself runs inside the runner's own job container, talking to
+# the host's docker daemon over a mounted socket (docker-outside-of-docker) -
+# `docker run -v "$CI_ROOT:/work"` would ask the *host* daemon to bind-mount a
+# path that only exists inside this job container, which fails. `docker cp`
+# instead copies the files by content, sidestepping the path mismatch.
+PW_VERSION="$(node -p "require('./src/admin/node_modules/@playwright/test/package.json').version")"
+log "Running Playwright admin e2e suite against $BASE_URL (playwright:v$PW_VERSION-noble)"
+
+PW_CONTAINER="$(docker create -e "E2E_BASE_URL=$BASE_URL" -w /work/src/admin "mcr.microsoft.com/playwright:v${PW_VERSION}-noble" npm run e2e)"
+docker cp "$CI_ROOT/." "$PW_CONTAINER:/work"
+pw_status=0
+docker start -a "$PW_CONTAINER" || pw_status=$?
+docker rm -f "$PW_CONTAINER" > /dev/null
+[[ "$pw_status" -eq 0 ]] || fail "Playwright e2e suite failed (exit $pw_status)"
+
+log "smoke-qa.sh complete - QA environment validated end to end"