Ekklesia/.github/workflows/ci.yml

name: CI

# The server runs Gitea 1.27.2 (docs/development/security.md, "What this server honours"):
#
#   * `concurrency:` is honoured: a newer push to a pull request cancels the run it supersedes, so a
#     cancelled run's own cleanup is never guaranteed (see the CI stack sweep in each stack job).
#   * `permissions:` is supported by 1.27 with Gitea's own scopes, but what the job token can do on
#     this server is not yet proven, so nothing here relies on it being read-only.
#   * `timeout-minutes:` is no longer listed as unsupported, but is not verified here either;
#     teardown relies on the stack sweep and the runner's own reaper.
#
# `persist-credentials: false` below is unaffected by any of this and is enforced by the action
# itself. See docs/development/security.md.

on:
  pull_request:
  push:
    branches:
      - main

permissions:
  contents: read

concurrency:
  group: ci-${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

jobs:
  verify:
    runs-on: ubuntu-latest
    timeout-minutes: 15
    steps:
      - name: Check out repository
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
          # `actions/checkout` defaults to leaving the job token in `.git/config`, where any later
          # step — including one a pull request added to this file — could read or reuse it. Nothing
          # in this workflow pushes, so the credential is never needed after checkout.
          persist-credentials: false

      - name: Set up Node.js
        uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
        with:
          node-version-file: .node-version

      - name: Activate pinned pnpm
        run: |
          corepack enable
          corepack prepare [email protected] --activate

      - name: Install dependencies
        run: pnpm install --frozen-lockfile

      - name: Verify repository
        run: pnpm ci:verify

  infrastructure-smoke:
    needs: verify
    runs-on: ubuntu-latest
    timeout-minutes: 10
    env:
      # Names this job's stacks, so its final step removes exactly these (scripts/lib/ci-project.sh).
      EKKLESIA_CI_JOB_KEY: infrastructure-smoke
    steps:
      - name: Check out repository
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
          # `actions/checkout` defaults to leaving the job token in `.git/config`, where any later
          # step — including one a pull request added to this file — could read or reuse it. Nothing
          # in this workflow pushes, so the credential is never needed after checkout.
          persist-credentials: false

      # A job the runner killed (cancelled, or past the runner's own time limit) never ran its
      # cleanup. Anything a CI job left on this shared daemon more than three hours ago goes first.
      - name: Sweep stale CI stacks from the runner
        run: bash scripts/ci-sweep.sh --stale

      - name: Prove the sweep removes exactly what it should, against real Docker resources
        run: bash scripts/test-ci-sweep.sh

      - name: Exercise PostgreSQL and S3 infrastructure
        run: bash scripts/smoke-infrastructure.sh

      # Whatever happened above, this job's own stacks do not outlive it.
      - name: Remove this job's CI stacks
        if: always()
        run: bash scripts/ci-sweep.sh --job

  migration-smoke:
    needs: verify
    runs-on: ubuntu-latest
    timeout-minutes: 15
    env:
      # Names this job's stacks, so its final step removes exactly these (scripts/lib/ci-project.sh).
      EKKLESIA_CI_JOB_KEY: migration-smoke
    steps:
      - name: Check out repository
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
          # `actions/checkout` defaults to leaving the job token in `.git/config`, where any later
          # step — including one a pull request added to this file — could read or reuse it. Nothing
          # in this workflow pushes, so the credential is never needed after checkout.
          persist-credentials: false

      # A job the runner killed (cancelled, or past the runner's own time limit) never ran its
      # cleanup. Anything a CI job left on this shared daemon more than three hours ago goes first.
      - name: Sweep stale CI stacks from the runner
        run: bash scripts/ci-sweep.sh --stale

      - name: Set up Node.js
        uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
        with:
          node-version-file: .node-version

      - name: Activate pinned pnpm
        run: |
          corepack enable
          corepack prepare [email protected] --activate

      - name: Install dependencies
        run: pnpm install --frozen-lockfile

      - name: Exercise transactional migrations against PostgreSQL
        run: pnpm test:migrations:integration

      # Whatever happened above, this job's own stacks do not outlive it.
      - name: Remove this job's CI stacks
        if: always()
        run: bash scripts/ci-sweep.sh --job

  package-release:
    needs: verify
    runs-on: ubuntu-latest
    timeout-minutes: 15
    # What the gate proved, for the job that consumes its artifact to check it received exactly that.
    outputs:
      version: ${{ steps.release.outputs.version }}
      revision: ${{ steps.release.outputs.revision }}
      manifest_sha256: ${{ steps.release.outputs.manifest_sha256 }}
    steps:
      - name: Check out repository
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
          # `actions/checkout` defaults to leaving the job token in `.git/config`, where any later
          # step — including one a pull request added to this file — could read or reuse it. Nothing
          # in this workflow pushes, so the credential is never needed after checkout.
          persist-credentials: false

      - name: Set up Node.js
        uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
        with:
          node-version-file: .node-version

      - name: Activate pinned pnpm
        run: |
          corepack enable
          corepack prepare [email protected] --activate

      - name: Install dependencies
        run: pnpm install --frozen-lockfile

      # ADR 0053's package release gate: build the public package train from this commit, build it
      # again from a second checkout and require byte-identical artifacts, then install exactly those
      # tarballs (offline, no registry) into a module outside this repository and prove it builds,
      # packages, and verifies. Nothing is published and no credential exists here. A future publish
      # job must depend on this job and publish only the artifacts it proved.
      - name: Build, reproduce, and consume the public package train
        run: pnpm test:package-release --keep package-train

      - name: Describe the proven release train
        id: release
        run: |
          release="$(find package-train -mindepth 1 -maxdepth 1 -type d)"
          manifest="$release/ekklesia-package-train.json"
          {
            echo "version=$(node -p 'JSON.parse(require("fs").readFileSync(process.argv[1], "utf8")).version' "$manifest")"
            echo "revision=$(node -p 'JSON.parse(require("fs").readFileSync(process.argv[1], "utf8")).source.revision' "$manifest")"
            echo "manifest_sha256=$(sha256sum "$manifest" | cut -d' ' -f1)"
          } >>"$GITHUB_OUTPUT"
          cat "$manifest"

      # The exact release artifacts, for the jobs that consume them, from a directory whose name does
      # not start with a dot: the artifact actions leave hidden files out. Version 3 of the artifact
      # actions: version 4 refuses any server that is not github.com, which this Gitea is not, while
      # Gitea implements the protocol version 3 uses. A consumer checks the files, not the wrapper.
      - name: Hand the release train on
        uses: actions/upload-artifact@c6a3b2bd78b3985e4b2f15397fec357f0fd808de # v3.2.2-node20
        with:
          name: package-train
          path: package-train/
          if-no-files-found: error
          retention-days: 7

  # The publication tool, given exactly the artifact the package-release job proved, against a
  # disposable Gitea 1.27.2 (scripts/test-package-publication.sh). Never the real registry, and no
  # credential: the disposable registry's accounts are created and discarded inside the job.
  package-publication:
    needs: package-release
    runs-on: ubuntu-latest
    timeout-minutes: 15
    env:
      # Names this job's stacks, so its final step removes exactly these (scripts/lib/ci-project.sh).
      EKKLESIA_CI_JOB_KEY: package-publication
    steps:
      - name: Check out repository
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
          persist-credentials: false

      # A job the runner killed (cancelled, or past the runner's own time limit) never ran its
      # cleanup. Anything a CI job left on this shared daemon more than three hours ago goes first.
      - name: Sweep stale CI stacks from the runner
        run: bash scripts/ci-sweep.sh --stale

      - name: Set up Node.js
        uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
        with:
          node-version-file: .node-version

      - name: Activate pinned pnpm
        run: |
          corepack enable
          corepack prepare [email protected] --activate

      - name: Install dependencies
        run: pnpm install --frozen-lockfile

      - name: Receive the release train the gate proved
        uses: actions/download-artifact@ad191675b41f6a5b46da9a048cb6893812da158b # v3.1.0-node20
        with:
          name: package-train
          path: package-train/

      - name: Check it is exactly that train
        env:
          EXPECTED_MANIFEST_SHA256: ${{ needs.package-release.outputs.manifest_sha256 }}
        run: |
          release="$(find package-train -mindepth 1 -maxdepth 1 -type d)"
          actual="$(sha256sum "$release/ekklesia-package-train.json" | cut -d' ' -f1)"
          if [[ -z "$EXPECTED_MANIFEST_SHA256" || "$actual" != "$EXPECTED_MANIFEST_SHA256" ]]; then
            echo "The received release manifest is not the one the package-release job proved." >&2
            exit 1
          fi
          (cd "$release" && sha256sum --check --strict SHA256SUMS)

      - name: Publish it to a disposable Gitea, and prove the round trip and recovery
        run: bash scripts/test-package-publication.sh --release "$(find package-train -mindepth 1 -maxdepth 1 -type d)"

      # Whatever happened above, this job's own stacks do not outlive it.
      - name: Remove this job's CI stacks
        if: always()
        run: bash scripts/ci-sweep.sh --job

  image-build:
    needs: verify
    runs-on: ubuntu-latest
    timeout-minutes: 20
    env:
      # Names this job's stacks, so its final step removes exactly these (scripts/lib/ci-project.sh).
      EKKLESIA_CI_JOB_KEY: image-build
      # Cache mounts in deploy/Dockerfile need BuildKit; this selects it even where the Docker CLI
      # has no buildx plugin.
      DOCKER_BUILDKIT: "1"
    steps:
      - name: Check out repository
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
          # `actions/checkout` defaults to leaving the job token in `.git/config`, where any later
          # step — including one a pull request added to this file — could read or reuse it. Nothing
          # in this workflow pushes, so the credential is never needed after checkout.
          persist-credentials: false

      # A job the runner killed (cancelled, or past the runner's own time limit) never ran its
      # cleanup. Anything a CI job left on this shared daemon more than three hours ago goes first.
      - name: Sweep stale CI stacks from the runner
        run: bash scripts/ci-sweep.sh --stale

      # The one generic platform image (ADR 0044, ADR 0048), proved rather than assumed: it contains no
      # module, no module template, and no module manifest; its module storage mount point is empty;
      # EKKLESIA_MODULES_DIR names it; and a checkout with a populated modules/ directory builds an
      # identical runtime file system. Build only — no registry login, no credentials, no push, no tag
      # anywhere but this runner's local image store. Tagged per run and attempt, and labeled like
      # every CI stack, so this job's final step removes the images even if a build is killed.
      - name: Build the generic platform image and prove it holds no modules
        run: bash scripts/check-image.sh

      # Whatever happened above, this job's own stacks do not outlive it.
      - name: Remove this job's CI stacks
        if: always()
        run: bash scripts/ci-sweep.sh --job

  deploy-smoke:
    needs: verify
    runs-on: ubuntu-latest
    timeout-minutes: 30
    # Both topologies (ADR 0045), each on the one generic image: same-origin, the default every
    # installation gets, and split, which installs a test-only module into module storage, removes
    # it, and restores it (ADR 0048), and drives a real headless browser across the two origins.
    strategy:
      fail-fast: false
      matrix:
        topology:
          - same-origin
          - split
    env:
      # Names this job's stacks, so its final step removes exactly these (scripts/lib/ci-project.sh).
      EKKLESIA_CI_JOB_KEY: deploy-smoke-${{ matrix.topology }}
      DOCKER_BUILDKIT: "1"
      EKKLESIA_SMOKE_TOPOLOGY: ${{ matrix.topology }}
    steps:
      - name: Check out repository
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
          # `actions/checkout` defaults to leaving the job token in `.git/config`, where any later
          # step — including one a pull request added to this file — could read or reuse it. Nothing
          # in this workflow pushes, so the credential is never needed after checkout.
          persist-credentials: false

      # A job the runner killed (cancelled, or past the runner's own time limit) never ran its
      # cleanup. Anything a CI job left on this shared daemon more than three hours ago goes first.
      - name: Sweep stale CI stacks from the runner
        run: bash scripts/ci-sweep.sh --stale

      # The reference deployment end to end (ADR 0044): the generic platform image, base + proxy with
      # Caddy's internal CA (test-only), an explicit migration apply against this job's own fresh
      # PostgreSQL, then the API, Dashboard, storage delivery, client-address chain, and verify-full
      # database TLS through the proxy. Every secret is generated for this run and discarded with
      # it; nothing is pushed, published, or logged in to.
      - name: Smoke-test the reference deployment (${{ matrix.topology }})
        run: bash scripts/smoke-deploy.sh

      # Whatever happened above, this job's own stacks do not outlive it.
      - name: Remove this job's CI stacks
        if: always()
        run: bash scripts/ci-sweep.sh --job