ci.yml

# ============================================================
# WorldScript Studio – CI/CD Pipeline
# security → quality (lint, i18n, tsgo, vitest+coverage) → build (+chunk budget, rollup analyze)
#   ├→ e2e | storybook (parallel) ; lighthouse after build
# deploy (main): needs ci-success → GitHub Pages
# ============================================================

name: CI / CD

on:
  push:
    branches: [main]
    tags: ['*']
  pull_request:
    branches: [main]
  workflow_dispatch:

permissions:
  contents: read

# QNBS-v3: cancel-in-progress only on PRs — cancelling a main-branch deploy mid-flight causes rollback confusion
concurrency:
  group: >-
    ci-${{ github.workflow }}-${{
      github.event_name == 'pull_request' && github.head_ref || github.ref_name
    }}
  cancel-in-progress: ${{ github.event_name == 'pull_request' }}

jobs:
  # ----------------------------------------------------------
  # -1. WORKFLOW POLICY GATE: runs before any job invokes the governed
  # ./.github/actions/setup composite (first used by "security" below), so a
  # PR that tampers with that composite — or any other workflow/action file —
  # is rejected before its own tampered code ever executes. Uses only
  # external, SHA-pinned actions directly, never the composite it gates.
  # ----------------------------------------------------------
  workflow-policy:
    name: 📜 Workflow Policy Gate
    runs-on: ubuntu-latest
    timeout-minutes: 10
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          persist-credentials: false
      # QNBS-v3: pnpm/setup instead of pnpm/action-setup, mirroring .github/actions/setup/action.yml (duplicated because this job predates the trust boundary it validates and can't use that composite).
      - uses: pnpm/setup@703c52620218391530e48b9e8870d5c0082e1b9b # v2.1.0
        with:
          version: 11.22.0
          install: false
      - uses: actions/setup-node@249970729cb0ef3589644e2896645e5dc5ba9c38 # v6.5.0
        with:
          node-version-file: .nvmrc
          cache: pnpm
      # QNBS-v3: no lifecycle/.pnpmfile.cjs hooks — install must not run repo code before the gate.
      - name: Install dependencies (no lifecycle scripts)
        run: pnpm install --frozen-lockfile --ignore-scripts --ignore-pnpmfile
      # QNBS-v3: node directly — pnpm run's own pre-run check rejects an --ignore-pnpmfile install.
      - name: Workflow-policy structural gate (permissions, needs graph, action pins)
        run: node scripts/workflow-policy-check.mjs

  # ----------------------------------------------------------
  # 0. SECURITY: pnpm audit + gitleaks + dependency review
  # ----------------------------------------------------------
  security:
    name: 🔒 Security Audit
    runs-on: ubuntu-latest
    timeout-minutes: 15
    needs: [workflow-policy]
    # QNBS-v3: dependency-review-action needs readable PR metadata — otherwise the job fails unnecessarily on external PRs/scopes.
    permissions:
      contents: read
      pull-requests: read
    steps:
      # QNBS-v3: fetch-depth:0 gives gitleaks access to parent commits (sha^) needed for PR diff scans; shallow clone causes "ambiguous argument" errors.
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          fetch-depth: 0
          persist-credentials: false
      - uses: ./.github/actions/setup

      # QNBS-v3: pnpm's audit client cannot decode force-gzipped responses from some npm CDN edge
      # nodes (ERR_PNPM_AUDIT_BAD_RESPONSE, deterministic per edge — retries are useless). The
      # enforced vulnerability gate in this job is the OSV scan below (covers BOTH pnpm-lock.yaml
      # and Cargo.lock with a superset DB); this step is advisory-only until pnpm fixes decoding.
      - name: pnpm audit (high + critical, advisory)
        continue-on-error: true
        run: |
          if pnpm audit --audit-level=high; then
            echo "pnpm audit: clean"
          else
            echo "::warning::pnpm audit failed (registry gzip-decoding bug or findings) — OSV scan below is the enforced gate"
            exit 1
          fi

      # QNBS-v3 (#60): vendored y-webrtc fork invariant guard (PBKDF2 600k, non-extractable keys,
      # DataChannel encryption, no dangling upstream dep) — fails the build on fork drift.
      - name: Vendor fork invariant guard
        run: pnpm run verify:vendor

      - name: OSV vulnerability scan
        uses: google/osv-scanner-action/osv-scanner-action@6e4298ebc4db23e847df9b2e2de2939d6f066c67 # v2.5.1
        with:
          scan-args: |-
            --config=src-tauri/osv-scanner.toml
            --lockfile=pnpm-lock.yaml
            --lockfile=src-tauri/Cargo.lock
            --lockfile=crates/Cargo.lock

      # QNBS-v3: gitleaks scans git history for secrets; GITHUB_TOKEN is enough for PR annotations.
      - name: Scan for leaked secrets (gitleaks)
        uses: gitleaks/gitleaks-action@e0c47f4f8be36e29cdc102c57e68cb5cbf0e8d1e # v3.0.0 — Node 24 compatible
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

      - name: Dependency review (PRs only)
        if: github.event_name == 'pull_request'
        uses: actions/dependency-review-action@a1d282b36b6f3519aa1f3fc636f609c47dddb294 # v5

  # ----------------------------------------------------------
  # 0b. CHANGE DETECTION: path-scopes Rust gates so a docs/frontend-only PR
  # doesn't pay for a Rust toolchain + apt-get(libgtk/libwebkit) build it can't
  # affect. Fails OPEN (tauri=true) on any ambiguity — base SHA missing/unreachable
  # — so a detection error runs the real gate instead of silently skipping it.
  # ----------------------------------------------------------
  changes:
    name: 📂 Detect changed paths
    runs-on: ubuntu-latest
    timeout-minutes: 5
    outputs:
      tauri: ${{ steps.filter.outputs.tauri }}
      crates: ${{ steps.filter.outputs.crates }}
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          fetch-depth: 0
          persist-credentials: false

      - name: Detect src-tauri / crates changes
        id: filter
        run: |
          if [ "${{ github.event_name }}" = "pull_request" ]; then
            BASE="${{ github.event.pull_request.base.sha }}"
          else
            BASE="${{ github.event.before }}"
          fi
          if [ -z "$BASE" ] || [ "$BASE" = "0000000000000000000000000000000000000000" ] || ! git cat-file -e "$BASE" 2>/dev/null; then
            echo "::notice::No usable base SHA to diff against — defaulting to tauri=true, crates=true (fail open)"
            echo "tauri=true" >> "$GITHUB_OUTPUT"
            echo "crates=true" >> "$GITHUB_OUTPUT"
            exit 0
          fi
          CHANGED=$(git diff --name-only "$BASE" "${{ github.sha }}")
          if grep -qE '^(src-tauri/|crates/|\.github/workflows/ci\.yml$)' <<< "$CHANGED"; then
            echo "tauri=true" >> "$GITHUB_OUTPUT"
          else
            echo "tauri=false" >> "$GITHUB_OUTPUT"
          fi
          if grep -qE '^(crates/|tests/fixtures/project-golden-masters/|\.github/workflows/ci\.yml$)' <<< "$CHANGED"; then
            echo "crates=true" >> "$GITHUB_OUTPUT"
          else
            echo "crates=false" >> "$GITHUB_OUTPUT"
          fi

  # ----------------------------------------------------------
  # PR SIZE GOVERNANCE: advisory below the absolute ceiling, blocking above it.
  # QNBS-v3: runs the BASE ref's checker copy — a PR touching it can't raise its own limits.
  # ----------------------------------------------------------
  pr-size:
    name: 📏 PR Size Governance
    if: github.event_name == 'pull_request'
    needs: [workflow-policy]
    runs-on: ubuntu-latest
    timeout-minutes: 5
    permissions:
      contents: read
      pull-requests: write
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          fetch-depth: 0
          persist-credentials: false

      - name: Evaluate PR size (base-ref checker)
        id: size
        env:
          GH_TOKEN: ${{ github.token }}
          BASE_SHA: ${{ github.event.pull_request.base.sha }}
          HEAD_SHA: ${{ github.event.pull_request.head.sha }}
        run: |
          set -euo pipefail
          mkdir -p /tmp/base-scripts
          # QNBS-v3: bootstrap fallback — the PR introducing this checker has no base-ref copy yet.
          if git show "$BASE_SHA:scripts/check-pr-size.mjs" > /tmp/base-scripts/check-pr-size.mjs 2>/dev/null \
             && git show "$BASE_SHA:scripts/ci-prepush-classifier.mjs" > /tmp/base-scripts/ci-prepush-classifier.mjs 2>/dev/null; then
            CHECKER=/tmp/base-scripts/check-pr-size.mjs
          else
            echo "::notice::check-pr-size.mjs not found on base ref (bootstrap PR) — using this PR's own copy this one time."
            CHECKER=scripts/check-pr-size.mjs
          fi
          set +e
          node "$CHECKER" "$BASE_SHA" "$HEAD_SHA" | tee /tmp/pr-size-report.txt
          exit_code=${PIPESTATUS[0]}
          set -e
          echo "exit_code=$exit_code" >> "$GITHUB_OUTPUT"

      # QNBS-v3: fork PRs get a read-only token — a failed comment must never fail this advisory job.
      # QNBS-v3: edits by owned comment ID (not --edit-last) so an intervening bot comment can't be overwritten.
      - name: Post PR size comment (advisory or blocking tiers only)
        continue-on-error: true
        env:
          GH_TOKEN: ${{ github.token }}
        run: |
          PR="${{ github.event.pull_request.number }}"
          REPO="${{ github.repository }}"
          MARKER="<!-- pr-size-governance -->"
          # QNBS-v3: --paginate + --jq filters each page separately — pipe to external jq -s to combine first.
          OWNED_ID=$(gh api --paginate "repos/$REPO/issues/$PR/comments" 2>/dev/null \
            | jq -s --arg marker "$MARKER" \
              '[.[][] | select(.user.login == "github-actions[bot]" and (.body | contains($marker)))] | last | .id // empty' \
            || echo "")
          if grep -qE 'within target' /tmp/pr-size-report.txt; then
            if [ -n "$OWNED_ID" ]; then
              printf '%s\nPR size is back within target — previous warning below is resolved.\n' "$MARKER" > /tmp/pr-size-resolved.txt
              gh api "repos/$REPO/issues/comments/$OWNED_ID" -X PATCH -F body=@/tmp/pr-size-resolved.txt || true
            fi
            exit 0
          fi
          { printf '%s\n' "$MARKER"; cat /tmp/pr-size-report.txt; } > /tmp/pr-size-marked.txt
          if [ -n "$OWNED_ID" ]; then
            gh api "repos/$REPO/issues/comments/$OWNED_ID" -X PATCH -F body=@/tmp/pr-size-marked.txt || true
          else
            gh pr comment "$PR" --body-file /tmp/pr-size-marked.txt || true
          fi

      # QNBS-v3: any nonzero exit (crash, OOM-kill, etc.) must fail, not just the literal '1' ceiling code.
      - name: Fail on a checker crash or an absolute-ceiling violation
        if: steps.size.outputs.exit_code != '0'
        run: exit 1

  signatures:
    name: 🔏 Verified Signatures
    runs-on: ubuntu-latest
    timeout-minutes: 10
    needs: [security]
    permissions:
      contents: read
      pull-requests: read
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          fetch-depth: 0
          persist-credentials: false
      - name: Verify every introduced GitHub commit and release tag
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: node scripts/signing/verify-github-signatures.mjs
      - name: Reject AI/session attribution in introduced commits
        run: |
          set -euo pipefail
          HEAD="${{ github.sha }}"
          # QNBS-v3: run the base-ref copy on PRs so this PR can't weaken the checker that grades it.
          CHECKER=scripts/check-commit-attribution.mjs
          if [ "${{ github.event_name }}" = "pull_request" ]; then
            BASE="${{ github.event.pull_request.base.sha }}"
            mkdir -p /tmp/base-scripts
            if git show "$BASE:scripts/check-commit-attribution.mjs" > /tmp/base-scripts/check-commit-attribution.mjs 2>/dev/null; then
              CHECKER=/tmp/base-scripts/check-commit-attribution.mjs
            else
              echo "::notice::check-commit-attribution.mjs not found on base ref (bootstrap PR) — using this PR's own copy this one time."
            fi
            node "$CHECKER" "$BASE" "$HEAD"
          elif [[ "$GITHUB_REF" == refs/tags/* ]]; then
            # QNBS-v3: github.sha is the dereferenced commit for a tag push; rev-parse the ref for the actual tag object so the annotation itself gets checked.
            TAG_SHA=$(git rev-parse "$GITHUB_REF")
            node "$CHECKER" --tag "$TAG_SHA"
          else
            BASE="${{ github.event.before }}"
            if [ -z "$BASE" ] || [ "$BASE" = "0000000000000000000000000000000000000000" ] || ! git cat-file -e "$BASE" 2>/dev/null; then
              echo "::error::No usable base SHA to diff against — cannot verify introduced commits for AI/session attribution. This is expected only on the repository's first-ever commit; review manually before merging."
              exit 1
            fi
            node "$CHECKER" "$BASE" "$HEAD"
          fi
          # QNBS-v3: PR title/body attribution is checked by the separate, cheap pr-text-attribution.yml (also runs on `edited`, which this workflow's trigger deliberately excludes).

  # ----------------------------------------------------------
  # 1. QUALITY GATE: Lint + Typecheck + Tests (parallel matrix)
  # ----------------------------------------------------------
  quality:
    name: 🔍 Quality Gate (Node ${{ matrix.node-version }})
    runs-on: ubuntu-latest
    timeout-minutes: 22
    needs: [security]
    strategy:
      fail-fast: false
      matrix:
        # QNBS-v3: Explicit versions — `node` (= 'current') breaks CI without code changes when Node N+1 is released.
        node-version: ['22', '24']
    steps:
      # QNBS-v3: fetch-tags — check-doc-metrics.mjs's stale-PLANNED-status check reads `git tag`
      # and silently no-ops without one; a default shallow checkout has no tags at all. fetch-depth:
      # 0 (full history) is additionally required by scanUnreleasedTruth's completeness check
      # (audit F-2): fetch-tags alone makes the tag ref resolvable but does not deepen the commit
      # graph, so `git log v<tag>..HEAD` could still fail or truncate on a shallow checkout — which
      # getPostReleaseCommitSubjects() would silently treat as "no history" and skip the entire
      # check, defeating it without ever reporting a failure.
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          persist-credentials: false
          fetch-tags: true
          fetch-depth: 0
      - uses: ./.github/actions/setup
        with:
          node-version: ${{ matrix.node-version }}

      - name: Lint (Biome)
        run: pnpm run lint

      - name: Suppression-debt ratchet (biome-ignore count)
        run: node scripts/check-suppressions.mjs

      - name: Token-audit ratchet (raw-color/inline-svg count)
        run: pnpm run token:audit

      - name: i18n key parity (vs en)
        run: pnpm run i18n:check

      - name: i18n translation quality floor
        run: pnpm run i18n:quality

      - name: Doc metrics drift gate (locale/key counts, stale PLANNED status)
        run: pnpm run docs:check

      - name: CSP source synchronization and parity
        run: pnpm run csp:verify

      - name: Native-readiness roadmap policy gate
        run: pnpm run native-readiness:check

      - name: Tauri plugin Rust/npm version-parity gate
        run: pnpm run tauri-plugins:check

      - name: Feature parity audit
        run: pnpm run parity:check

      - name: DesktopPlatform import-boundary gate (Wave 1)
        run: pnpm run guardrail:desktop-imports

      - name: Typecheck (tsgo)
        run: npx tsgo --project tsconfig.tsgo.json --noEmit --checkers 4

      # QNBS-v3: no retry — a first-attempt failure remains visible instead of being masked as a pass.
      # --no-experimental-webstorage disables Node 24's native Web Storage API for deterministic setup.
      # QNBS-v3: both outputFile paths explicit in this one invocation — never relying on vitest.config.ts's own reporter-tuple default, which could silently drift independently of this exact command.
      - name: Unit tests (Vitest, no retry)
        run: pnpm exec vitest run --coverage --reporter=json --reporter=junit --outputFile.json=test-results.json --outputFile.junit=reports/junit.xml
        env:
          NODE_OPTIONS: "--no-experimental-webstorage --max-old-space-size=4096"
          CI: "true"

      - name: Coverage ratchet check (informational, non-blocking)
        if: always()
        continue-on-error: true
        run: node scripts/check-coverage-ratchet.mjs

      - name: Upload Vitest JSON results
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: vitest-results-node${{ matrix.node-version }}
          path: test-results.json
          if-no-files-found: ignore
          retention-days: 7

      - name: Upload Vitest JUnit results
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: vitest-junit-node${{ matrix.node-version }}
          path: reports/junit.xml
          if-no-files-found: ignore
          retention-days: 7

      - name: Upload coverage report
        # QNBS-v3: matrix value stays literal 'lts/*' — one upload per workflow suffices (Codecov optional).
        if: always() && matrix.node-version == '22'
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: coverage-report
          path: coverage/
          if-no-files-found: warn
          retention-days: 7

      # QNBS-v3: fail_ci_if_error: false — forks without CODECOV_TOKEN; coverage remains in artifact regardless.
      - name: Publish coverage to Codecov
        if: always() && matrix.node-version == '22'
        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f # v7.0.0
        with:
          token: ${{ secrets.CODECOV_TOKEN }}
          files: coverage/lcov.info
          flags: unit
          fail_ci_if_error: false

      # QNBS-v3: !cancelled() (not the default success()) — a real test failure must still upload results, only a cancelled run skips it.
      - name: Publish unit test results to Codecov
        if: ${{ !cancelled() }}
        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f # v7.0.0
        with:
          token: ${{ secrets.CODECOV_TOKEN }}
          report_type: test_results
          files: reports/junit.xml
          disable_search: true
          flags: unit-node${{ matrix.node-version }}
          name: unit-node${{ matrix.node-version }}
          fail_ci_if_error: false

  # ----------------------------------------------------------
  # 1b. RUST-TAURI: Tauri Rust Gate (fmt/check/clippy/test), path-scoped via `changes`
  # ----------------------------------------------------------
  rust-tauri:
    name: 🦀 Tauri Rust Gate
    runs-on: ubuntu-latest
    timeout-minutes: 20
    needs: [security, changes]
    # QNBS-v3: skips for PRs that don't touch src-tauri/** — ci-success treats 'skipped' as pass for this job only
    if: needs.changes.outputs.tauri == 'true'
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          persist-credentials: false
      - uses: dtolnay/rust-toolchain@29eef336d9b2848a0b548edc03f92a220660cdb8 # stable
        with:
          toolchain: stable
          components: rustfmt, clippy
      # QNBS-v3: no cargo caching existed before — fmt/check/clippy/test each rebuilt the full
      # dependency tree from scratch every run. Keyed on Cargo.lock + rustc version, so a toolchain
      # bump or lockfile change invalidates cleanly rather than reusing stale artifacts.
      - uses: Swatinem/rust-cache@6323deb102c322ba6fcbdcafc7e3dddab59af2b6 # v2.9.2
        with:
          workspaces: src-tauri -> target
      # QNBS-v3: /var/cache/apt/archives is root-owned by default — actions/cache's restore
      # (tar extraction) runs as the unprivileged runner user and fails with "Permission denied"
      # on every file without this, silently degrading to a cache miss every time (real CI
      # evidence: "Cache hit" followed immediately by "Cache not found" from the failed tar).
      - name: Make apt archive cache dir writable by the runner user
        run: sudo mkdir -p /var/cache/apt/archives/partial && sudo chown -R runner:runner /var/cache/apt/archives && sudo chmod -R 755 /var/cache/apt/archives

      - name: Cache apt packages (Tauri Linux build deps)
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: /var/cache/apt/archives/*.deb
          # QNBS-v3: bump the -v suffix if the apt-get install package list below changes
          key: apt-tauri-linux-deps-v1
      - name: Install Linux Tauri build dependencies
        run: |
          sudo apt-get update
          sudo apt-get install -y libgtk-3-dev libwebkit2gtk-4.1-dev libsoup-3.0-dev libayatana-appindicator3-dev librsvg2-dev patchelf
      - name: Rust format check
        working-directory: src-tauri
        run: cargo fmt --check
      - name: Rust compile check
        working-directory: src-tauri
        run: cargo check --locked
      - name: Rust clippy
        working-directory: src-tauri
        run: cargo clippy --locked --all-targets -- -D warnings
      - name: Rust tests
        working-directory: src-tauri
        run: cargo test --locked

  # ----------------------------------------------------------
  # 1c. CORE-RUST: worldscript-project crate Gate (fmt/check/clippy/test), path-scoped via `changes`
  # ----------------------------------------------------------
  core-rust:
    name: 🧩 Core Rust Gate
    runs-on: ubuntu-latest
    timeout-minutes: 15
    needs: [security, changes]
    if: needs.changes.outputs.crates == 'true'
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          persist-credentials: false
      - uses: dtolnay/rust-toolchain@29eef336d9b2848a0b548edc03f92a220660cdb8 # stable
        with:
          toolchain: stable
          components: rustfmt, clippy
      - uses: Swatinem/rust-cache@6323deb102c322ba6fcbdcafc7e3dddab59af2b6 # v2.9.2
        with:
          workspaces: crates -> target
      - name: Rust format check
        working-directory: crates
        run: cargo fmt --check
      - name: Rust compile check
        working-directory: crates
        run: cargo check --locked
      - name: Rust clippy
        working-directory: crates
        run: cargo clippy --locked --all-targets -- -D warnings
      - name: Rust tests
        working-directory: crates
        run: cargo test --locked

  # ----------------------------------------------------------
  # 2. BUILD: Production build + Pages artifact upload
  # QNBS-v3: Stryker mutation removed from the PR/CI pipeline (2026-06-02) — it added noise as a
  #          flaky, non-gating check. Mutation now runs ONLY via the manual `.github/workflows/
  #          mutation.yml` (workflow_dispatch). To be re-integrated in a later iteration.
  # ----------------------------------------------------------
  build:
    name: 🏗️ Build
    runs-on: ubuntu-latest
    timeout-minutes: 20
    needs: [quality]
    # QNBS-v3: attestations + id-token needed for SLSA build provenance on main.
    permissions:
      contents: read
      attestations: write
      id-token: write
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          persist-credentials: false
      - uses: ./.github/actions/setup

      - name: Build application
        run: pnpm run build
        env:
          NODE_ENV: production

      # QNBS-v3: caches the apt .deb archives `--with-deps` fetches (fonts, libx11, etc.) — shared
      # key across every `playwright install --with-deps` call site in this workflow, since they all
      # install the same system packages on the same runner image. actions/cache/save on a fresh
      # write here also warms the cache for e2e/e2e-deep/storybook/vrt below.
      # QNBS-v3: /var/cache/apt/archives is root-owned by default — actions/cache's restore
      # (tar extraction) runs as the unprivileged runner user and fails with "Permission denied"
      # on every file without this, silently degrading to a cache miss every time (real CI
      # evidence: "Cache hit" followed immediately by "Cache not found" from the failed tar).
      - name: Make apt archive cache dir writable by the runner user
        run: sudo mkdir -p /var/cache/apt/archives/partial && sudo chown -R runner:runner /var/cache/apt/archives && sudo chmod -R 755 /var/cache/apt/archives

      - name: Cache apt packages (Playwright system deps)
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: /var/cache/apt/archives/*.deb
          key: apt-playwright-chromium-deps-v1

      # QNBS-v3: The E2E suite runs `vite dev`, so prod-only rolldown bundling crashes (e.g. the
      # zod DCE "init_locales is not defined" blank screen) never surface there. This guard loads
      # the real production bundle in headless Chromium and fails if React does not mount.
      - name: Smoke-test production build (mounts, no runtime crash)
        run: |
          pnpm exec playwright install --with-deps chromium-headless-shell
          pnpm run smoke:prod

      - name: Bundle chunk size budget
        run: pnpm run bundle:budget

      # QNBS-v3: token scoped to this exact step only, never job/workflow-wide — Codecov Bundle Analysis needs it inside this one Vite build, nothing else does.
      - name: Bundle analysis (rollup visualizer + Codecov)
        run: pnpm run analyze
        env:
          NODE_ENV: production
          CODECOV_BUNDLE_ANALYSIS: 'true'
          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
          # QNBS-v3: PR head SHA (not GitHub's synthetic merge commit) so Bundle Analysis attributes to the same commit reviewers see; falls back to github.sha on main pushes.
          CODECOV_BUNDLE_SHA: ${{ github.event.pull_request.head.sha || github.sha }}

      - name: Upload bundle analysis HTML
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: bundle-analysis
          path: dist/bundle-analysis.html
          retention-days: 14

      - name: Upload build artifact
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: dist
          path: ./dist
          # QNBS-v3: 3 days — dist is only needed by lighthouse + deploy jobs in the same run
          retention-days: 3

      # QNBS-v3: SLSA provenance attestation — ties the dist artifact to this exact
      # workflow run and commit SHA; verifiable with `gh attestation verify`. Excludes
      # woff/woff2 — self-hosted CJK fonts (@fontsource noto-sans-jp/kr/sc) ship ~2.2k
      # unicode-range subset files, pushing dist past the action's 1024-subject cap.
      # Fonts are vendored static assets, not application code, so attesting the app
      # bundle (JS/CSS/HTML/JSON/wasm/etc.) without them still covers what matters.
      - name: Attest build provenance
        if: github.ref == 'refs/heads/main' && github.event_name != 'pull_request'
        uses: actions/attest-build-provenance@4d101475d8b20a2381f78447822ac1eab6504dd8 # v4.2.2
        with:
          subject-path: |
            dist/**
            !dist/**/*.woff
            !dist/**/*.woff2

      - name: Setup Pages
        if: github.ref == 'refs/heads/main' && github.event_name != 'pull_request'
        uses: actions/configure-pages@45bfe0192ca1faeb007ade9deae92b16b8254a0d # v6

      - name: Upload Pages artifact
        if: github.ref == 'refs/heads/main' && github.event_name != 'pull_request'
        uses: actions/upload-pages-artifact@fc324d3547104276b827a68afc52ff2a11cc49c9 # v5.0.0
        with:
          path: ./dist

  # ----------------------------------------------------------
  # 3. CI SUCCESS: single required-status aggregator
  #    (security + signatures + quality + changes + rust-tauri + core-rust + build + e2e + vrt)
  # ----------------------------------------------------------
  ci-success:
    name: ✅ CI Success
    runs-on: ubuntu-latest
    timeout-minutes: 5
    # QNBS-v3: Every unconditional job is either required here or explicitly advisory at job level.
    needs:
      [
        workflow-policy,
        pr-size,
        security,
        signatures,
        quality,
        changes,
        rust-tauri,
        core-rust,
        build,
        e2e,
        browser-quality,
      ]
    if: always()
    steps:
      - name: Verify all required jobs succeeded
        run: |
          FAIL=0
          [ "${{ needs.workflow-policy.result }}" = "success" ] || FAIL=1
          # QNBS-v3: pr-size's own if: matches this event check — unconditional skip-tolerance would let a tampered if: pass.
          if [ "${{ github.event_name }}" = "pull_request" ]; then
            [ "${{ needs.pr-size.result }}" = "success" ] || FAIL=1
          elif [ "${{ needs.pr-size.result }}" != "success" ] && [ "${{ needs.pr-size.result }}" != "skipped" ]; then
            FAIL=1
          fi
          [ "${{ needs.security.result }}" = "success" ] || FAIL=1
          [ "${{ needs.signatures.result }}" = "success" ] || FAIL=1
          [ "${{ needs.quality.result }}" = "success" ] || FAIL=1
          [ "${{ needs.changes.result }}" = "success" ] || FAIL=1
          if [ "${{ needs.rust-tauri.result }}" != "success" ] && [ "${{ needs.rust-tauri.result }}" != "skipped" ]; then
            FAIL=1
          fi
          if [ "${{ needs.core-rust.result }}" != "success" ] && [ "${{ needs.core-rust.result }}" != "skipped" ]; then
            FAIL=1
          fi
          [ "${{ needs.build.result }}" = "success" ] || FAIL=1
          [ "${{ needs.e2e.result }}" = "success" ] || FAIL=1
          [ "${{ needs.browser-quality.result }}" = "success" ] || FAIL=1
          if [ "$FAIL" = "1" ]; then
            echo "One or more required jobs did not succeed:"
            echo "  workflow-policy: ${{ needs.workflow-policy.result }}"
            echo "  pr-size: ${{ needs.pr-size.result }} (skipped = OK, not a pull_request event)"
            echo "  security: ${{ needs.security.result }}"
            echo "  signatures: ${{ needs.signatures.result }}"
            echo "  quality:  ${{ needs.quality.result }}"
            echo "  changes:  ${{ needs.changes.result }}"
            echo "  rust-tauri: ${{ needs.rust-tauri.result }} (skipped = OK, src-tauri untouched)"
            echo "  core-rust: ${{ needs.core-rust.result }} (skipped = OK, crates/ untouched)"
            echo "  build:    ${{ needs.build.result }}"
            echo "  e2e:      ${{ needs.e2e.result }}"
            echo "  browser-quality: ${{ needs.browser-quality.result }}"
            exit 1
          fi
          echo "All required jobs succeeded (or were legitimately skipped)."

  # ----------------------------------------------------------
  # 4. DEPLOY: GitHub Pages (only on main push)
  # ----------------------------------------------------------
  deploy:
    name: 🚀 Deploy to GitHub Pages
    runs-on: ubuntu-latest
    timeout-minutes: 10
    needs: [ci-success]
    # QNBS-v3 (#522): always()+!cancelled() overrides GHA's default skip-propagation from pr-size/rust-tauri/core-rust being legitimately skipped on push, without publishing from a cancelled run — ci-success.result is the real gate.
    if: >-
      ${{
        always() &&
        !cancelled() &&
        github.ref == 'refs/heads/main' &&
        github.event_name != 'pull_request' &&
        needs.ci-success.result == 'success'
      }}
    permissions:
      contents: read
      pages: write
      id-token: write
    environment:
      name: github-pages
      url: ${{ steps.deployment.outputs.page_url }}
    steps:
      - name: Deploy to GitHub Pages
        id: deployment
        uses: actions/deploy-pages@cd2ce8fcbc39b97be8ca5fce6e763baed58fa128 # v5.0.0

  # ----------------------------------------------------------
  # 5. E2E: Playwright (after quality gate, parallel with build)
  # ----------------------------------------------------------
  e2e:
    name: 🎭 E2E Tests (Playwright)
    runs-on: ubuntu-latest
    # QNBS-v3: 29 tests × 2 browser-projects × retries:2 sequential → can exceed 30 min; raised to 50.
    timeout-minutes: 50
    needs: [quality]
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          persist-credentials: false
      - uses: ./.github/actions/setup

      # QNBS-v3: cache Playwright browsers by version so re-runs don't re-download ~400 MB each time.
      - name: Cache Playwright browsers
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: ~/.cache/ms-playwright
          key: playwright-${{ runner.os }}-${{ hashFiles('**/pnpm-lock.yaml') }}

      # QNBS-v3: /var/cache/apt/archives is root-owned by default — actions/cache's restore
      # (tar extraction) runs as the unprivileged runner user and fails with "Permission denied"
      # on every file without this, silently degrading to a cache miss every time (real CI
      # evidence: "Cache hit" followed immediately by "Cache not found" from the failed tar).
      - name: Make apt archive cache dir writable by the runner user
        run: sudo mkdir -p /var/cache/apt/archives/partial && sudo chown -R runner:runner /var/cache/apt/archives && sudo chmod -R 755 /var/cache/apt/archives

      - name: Cache apt packages (Playwright system deps)
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: /var/cache/apt/archives/*.deb
          key: apt-playwright-chromium-deps-v1

      - name: Install Playwright browsers
        run: pnpm exec playwright install --with-deps chromium

      - name: Run E2E tests
        run: pnpm run test:e2e
        env:
          CI: 'true'
          # QNBS-v3: Skip VRT in E2E job — visual regression is handled by the dedicated
          # vrt job which serves the production build and owns the baseline PNGs.
          PLAYWRIGHT_SKIP_VRT: 'true'

      - name: Upload Playwright HTML report
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: playwright-report
          path: tests/e2e/html-report/
          if-no-files-found: warn
          retention-days: 7

      # QNBS-v3: JUnit output from playwright.config.ts reporter — enables per-test annotation in GitHub PR checks
      - name: Upload E2E JUnit results
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: e2e-junit
          path: tests/e2e/results/junit.xml
          if-no-files-found: warn
          retention-days: 7

      - name: Publish E2E test results to Codecov
        if: ${{ !cancelled() }}
        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f # v7.0.0
        with:
          token: ${{ secrets.CODECOV_TOKEN }}
          report_type: test_results
          files: tests/e2e/results/junit.xml
          disable_search: true
          flags: e2e
          name: playwright-e2e
          fail_ci_if_error: false

  # ----------------------------------------------------------
  # 6a. E2E DEEP COVERAGE: Feature-flag matrix + error paths (non-blocking)
  # QNBS-v3: Separate from the required `e2e` gate. Runs tests/e2e/deep/ with
  # explicit flag combinations that were previously invisible (coverage illusion).
  # continue-on-error:true so deep failures inform but never block a merge.
  # ----------------------------------------------------------
  e2e-deep:
    name: 🔬 E2E Deep Coverage (feature-flag matrix)
    runs-on: ubuntu-latest
    timeout-minutes: 30
    needs: [quality]
    continue-on-error: true
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          persist-credentials: false
      - uses: ./.github/actions/setup

      - name: Cache Playwright browsers
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: ~/.cache/ms-playwright
          key: playwright-${{ runner.os }}-${{ hashFiles('**/pnpm-lock.yaml') }}

      # QNBS-v3: /var/cache/apt/archives is root-owned by default — actions/cache's restore
      # (tar extraction) runs as the unprivileged runner user and fails with "Permission denied"
      # on every file without this, silently degrading to a cache miss every time (real CI
      # evidence: "Cache hit" followed immediately by "Cache not found" from the failed tar).
      - name: Make apt archive cache dir writable by the runner user
        run: sudo mkdir -p /var/cache/apt/archives/partial && sudo chown -R runner:runner /var/cache/apt/archives && sudo chmod -R 755 /var/cache/apt/archives

      - name: Cache apt packages (Playwright system deps)
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: /var/cache/apt/archives/*.deb
          key: apt-playwright-chromium-deps-v1

      - name: Install Playwright browsers
        run: pnpm exec playwright install --with-deps chromium

      - name: Run deep E2E coverage
        run: pnpm run test:e2e:deep
        env:
          CI: 'true'

      - name: Upload deep E2E report
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: e2e-deep-report
          path: tests/e2e/html-report/
          if-no-files-found: warn
          retention-days: 7

      # QNBS-v3: same playwright.config.ts reporter as the required e2e job, produced independently on this runner — previously never uploaded.
      - name: Upload deep E2E JUnit results
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: e2e-deep-junit
          path: tests/e2e/results/junit.xml
          if-no-files-found: warn
          retention-days: 7

      - name: Publish deep E2E test results to Codecov
        if: ${{ !cancelled() }}
        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f # v7.0.0
        with:
          token: ${{ secrets.CODECOV_TOKEN }}
          report_type: test_results
          files: tests/e2e/results/junit.xml
          disable_search: true
          flags: e2e-deep
          name: e2e-deep
          fail_ci_if_error: false

  # ----------------------------------------------------------
  # 6. BROWSER QUALITY: VRT + Lighthouse (after build)
  # ----------------------------------------------------------
  # QNBS-v3: consolidated from separate vrt/lighthouse jobs (Wave 2 CI-performance) — both already ran off the critical path; sharing checkout/setup/artifact-download once saves runner-minutes, measured to fit inside this baseline's critical-path slack, not a universal zero-wall-clock guarantee.
  browser-quality:
    name: 🔦🖼 Browser Quality (VRT + Lighthouse)
    runs-on: ubuntu-latest
    # QNBS-v3: union of the former 15 min (vrt) + 25 min (lighthouse) timeouts, now sequential in one job.
    timeout-minutes: 35
    needs: [build]
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          persist-credentials: false
      - uses: ./.github/actions/setup

      - name: Download build artifact
        uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8
        with:
          name: dist
          path: dist/

      # ---- VRT phase ----
      # QNBS-v3: /var/cache/apt/archives is root-owned by default — actions/cache's restore
      # (tar extraction) runs as the unprivileged runner user and fails with "Permission denied"
      # on every file without this, silently degrading to a cache miss every time (real CI
      # evidence: "Cache hit" followed immediately by "Cache not found" from the failed tar).
      - name: Make apt archive cache dir writable by the runner user
        run: sudo mkdir -p /var/cache/apt/archives/partial && sudo chown -R runner:runner /var/cache/apt/archives && sudo chmod -R 755 /var/cache/apt/archives

      - name: Cache apt packages (Playwright system deps)
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: /var/cache/apt/archives/*.deb
          key: apt-playwright-chromium-deps-v1

      - name: Install Playwright browsers
        run: pnpm exec playwright install --with-deps chromium

      - name: Serve dist + run VRT
        env:
          CI: true
        run: |
          # QNBS-v3: The GitHub Pages build has base=/WorldScript-Studio/. Mirror that structure
          # so http-server serves assets at the expected paths matching playwright.config baseURL.
          mkdir -p serve_root/WorldScript-Studio
          cp -r dist/. serve_root/WorldScript-Studio/
          pnpm exec http-server serve_root -p 3000 -s &
          pnpm exec wait-on http://127.0.0.1:3000 --timeout 30000
          pnpm run test:vrt

      - name: Upload VRT screenshots
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: vrt-screenshots
          # QNBS-v3: upload both baselines and diff outputs so PR reviewers can compare visually — playwright.config.ts sets no outputDir, so Playwright's default (repo-root test-results/, not tests/e2e/test-results/) is where actual/diff failure images actually land.
          path: |
            tests/e2e/visual-regression.spec.ts-snapshots/
            test-results/
          retention-days: 7

      # QNBS-v3: same playwright.config.ts reporter as the required e2e job, produced independently on this runner — previously never uploaded.
      - name: Upload VRT JUnit results
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: vrt-junit
          path: tests/e2e/results/junit.xml
          if-no-files-found: warn
          retention-days: 7

      - name: Publish VRT test results to Codecov
        if: ${{ !cancelled() }}
        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f # v7.0.0
        with:
          token: ${{ secrets.CODECOV_TOKEN }}
          report_type: test_results
          files: tests/e2e/results/junit.xml
          disable_search: true
          flags: vrt
          name: vrt
          fail_ci_if_error: false

      # ---- Lighthouse phase ----
      # QNBS-v3: !cancelled(), not the step default success() — VRT failing must not skip Lighthouse; each phase is an independent diagnostic gate that must produce its own evidence, matching what two separate jobs would have done.
      # QNBS-v3: no --assert.exitCode flag — let .lighthouserc.cjs assertions drive the
      # exit code; error-level assertions (CLS, accessibility) will correctly fail this job.
      - name: Run Lighthouse CI (mobile)
        if: ${{ !cancelled() }}
        run: pnpm exec lhci autorun
        env:
          LHCI_GITHUB_APP_TOKEN: ${{ secrets.LHCI_GITHUB_APP_TOKEN }}

      # QNBS-v3: !cancelled() so desktop still runs even if VRT or mobile Lighthouse failed; pre-existing continue-on-error (unchanged by this consolidation) means its accessibility/CLS error assertions can still fail this step internally without failing the job — promotion to blocking stays gated on the 5-consecutive-main-run exit criterion in docs/CI.md.
      - name: Run Lighthouse CI (desktop)
        if: ${{ !cancelled() }}
        run: pnpm exec lhci autorun --config=.lighthouserc.desktop.cjs
        continue-on-error: true
        env:
          LHCI_GITHUB_APP_TOKEN: ${{ secrets.LHCI_GITHUB_APP_TOKEN }}

      - name: Upload Lighthouse report
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: lighthouse-report
          path: .lighthouseci/
          if-no-files-found: warn
          # QNBS-v3: align with other report artifacts (7 days); 14 was inconsistent
          retention-days: 7

  # ----------------------------------------------------------
  # 7. STORYBOOK: Build + artifact (parallel with build)
  # ----------------------------------------------------------
  storybook:
    name: 📖 Storybook
    runs-on: ubuntu-latest
    # QNBS-v3: Cloud-first — Storybook build + test-runner runs only in CI (not local).
    #          25 min = build (~8 min) + server start + test-runner retries + screenshot upload.
    timeout-minutes: 25
    # QNBS-v3: Storybook remains advisory until ten genuine first-attempt runs establish stability.
    continue-on-error: true
    needs: [quality]
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          persist-credentials: false
      - uses: ./.github/actions/setup

      # QNBS-v3: Cache Playwright browsers between runs — saves ~60 s install time.
      - name: Cache Playwright browsers
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: ~/.cache/ms-playwright
          key: playwright-chromium-${{ runner.os }}-${{ hashFiles('pnpm-lock.yaml') }}
          restore-keys: playwright-chromium-${{ runner.os }}-

      - name: Build Storybook
        run: pnpm exec storybook build --output-dir storybook-static

      # QNBS-v3: /var/cache/apt/archives is root-owned by default — actions/cache's restore
      # (tar extraction) runs as the unprivileged runner user and fails with "Permission denied"
      # on every file without this, silently degrading to a cache miss every time (real CI
      # evidence: "Cache hit" followed immediately by "Cache not found" from the failed tar).
      - name: Make apt archive cache dir writable by the runner user
        run: sudo mkdir -p /var/cache/apt/archives/partial && sudo chown -R runner:runner /var/cache/apt/archives && sudo chmod -R 755 /var/cache/apt/archives

      - name: Cache apt packages (Playwright system deps)
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: /var/cache/apt/archives/*.deb
          key: apt-playwright-chromium-deps-v1

      # QNBS-v3: Chromium-only; test-runner v0.24+ is async-compatible with Storybook v10.
      - name: Install Playwright browsers (test-runner)
        run: pnpm exec playwright install --with-deps chromium

      - name: Serve Storybook + Run test-runner (cloud-optimised)
        # QNBS-v3 (F-12, 2026-07-29): the command below was invoking three flags this
        # @storybook/test-runner@^0.24.4 CLI does not support (`--max-workers` — wrong case, should
        # be `--maxWorkers`; `--retries` and `--screenshot-on-failure` don't exist in this version
        # at all). Every prior CI run failed on the FIRST unknown option before any story ever ran,
        # printed --help text, and `|| true` silently turned that into a green step — this gate has
        # produced zero real signal since it was added. Fixed to only pass supported flags.
        # QNBS-v3 (CodeRabbit, 2026-07-29): `|| true` swallowed the exit code so this step showed
        # green even when it truly failed — literally how the bug above stayed invisible. The
        # job-level advisory policy keeps the failed step visible without making it a merge blocker.
        env:
          CI: true
          DEBUG: pw:browser*
          # QNBS-v3: --junit's underlying jest-junit defaults to <cwd>/junit.xml, outside every path this job already uploads — pin it into test-results/ instead of guessing at the default.
          JEST_JUNIT_OUTPUT_FILE: test-results/storybook-junit.xml
        run: |
          npx http-server storybook-static -p 6006 -s &
          npx wait-on http://localhost:6006 --timeout 90000
          pnpm exec test-storybook --url http://localhost:6006 \
            --maxWorkers=2 \
            --junit

      # QNBS-v3: always() — upload even on failure; artifacts are the primary debug tool on low-end
      #          local hardware where storybook cannot run. Use the Storybook Debug workflow for
      #          interactive investigation (see .github/workflows/storybook-debug.yml).
      - name: Upload Storybook build + test results
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: storybook-build-and-test-results
          path: |
            storybook-static/
            test-results/
            **/screenshots/
          retention-days: 7

      - name: Publish Storybook test results to Codecov
        if: ${{ !cancelled() }}
        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f # v7.0.0
        with:
          token: ${{ secrets.CODECOV_TOKEN }}
          report_type: test_results
          files: test-results/storybook-junit.xml
          disable_search: true
          flags: storybook
          name: storybook
          fail_ci_if_error: false