opencode-mobile/.github/workflows/activation-e2e.yml

name: Activation E2E (Maestro)

# Deterministic regression coverage for the activation flow (first open ->
# telemetry consent -> server URL entry -> connect -> send first message),
# including the connect-time-401 negative case tied to the 0%-7-day-retention
# / GitHub issue #76 investigation.
#
# Also runs (non-blocking — see scripts/run-e2e-flows.sh's CORE_FLOWS vs
# NEWER_FLOWS split, and issue #104) newer surfaces merged after the initial
# activation suite: DirectoryBrowserSheet's server-folder picker, the
# directory-less "all sessions across all projects" list (+ the #46/#48
# open-across-project regression), VariantPicker's reasoning-effort chip, and
# DiffView/CodeBlock horizontal scroll. See .maestro/flows/directory-picker.yaml,
# all-sessions.yaml, variant-picker.yaml, diff-scroll.yaml. These never once
# ran to completion in CI (they sat behind the activation flows' stale
# assertions fixed by PR #102) and need their own hardening — tracked in #104.
#
# Runs against tests/fixtures/mock-opencode-server.ts (a small dependency-free
# HTTP+SSE stub matching the REAL client protocol read from src/lib/sdk.ts —
# NOT a live opencode server, NOT a WebSocket), so the suite is fast and fully
# self-contained: no external server, no LLM provider, no network flakiness.
#
# This is intentionally separate from cua-smoke.yml (the existing
# vision-driven CUA harness): that one needs a live opencode server + an Azure
# LLM and is exploratory/non-deterministic by design, so it isn't suited to
# tight regression assertions like "a 401 must show a visible error."
#
# activation-positive.yaml does NOT assert on the SSE-streamed reply — see
# the comment at the top of that file (issue #90 mode B / PR #102) for why:
# this Android-emulator + Node-mock + adb-reverse combination reliably
# stalls a long-lived SSE connection after its first chunk regardless of
# client transport, a CI-harness limitation with no evidence it affects real
# devices against a real server, so asserting on it here would be asserting
# on the harness rather than the app.

on:
  push:
    branches: [main]
    paths:
      - "app/**"
      - "src/**"
      - "tests/fixtures/**"
      - ".maestro/**"
      - ".github/workflows/activation-e2e.yml"
  pull_request:
    branches: [main]
    paths:
      - "app/**"
      - "src/**"
      - "tests/fixtures/**"
      - ".maestro/**"
      - ".github/workflows/activation-e2e.yml"
  workflow_dispatch: {}

jobs:
  activation-e2e:
    runs-on: ubuntu-latest
    # Same npm install + expo prebuild + assembleRelease + emulator pipeline as
    # cua-smoke.yml, which budgets 60 min (emulator-boot-timeout alone is 10 min).
    # Typical runs finish well under this; the ceiling just avoids flaky kills
    # on cold Gradle caches.
    timeout-minutes: 60
    steps:
      - uses: actions/checkout@v6

      - uses: actions/setup-node@v6
        with:
          node-version: 24
          cache: npm

      - uses: actions/setup-java@v5
        with:
          distribution: temurin
          java-version: 17

      - name: Setup Android SDK
        uses: android-actions/setup-android@v4

      - name: Add emulator to PATH
        run: echo "$ANDROID_HOME/emulator" >> $GITHUB_PATH

      - name: Enable KVM
        run: |
          echo 'KERNEL=="kvm", GROUP="kvm", MODE="0666", OPTIONS+="static_node=kvm"' | sudo tee /etc/udev/rules.d/99-kvm4all.rules
          sudo udevadm control --reload-rules
          sudo udevadm trigger --name-match=kvm

      - name: Cache Gradle
        uses: actions/cache@v5
        with:
          path: |
            ~/.gradle/caches
            ~/.gradle/wrapper
            android/.gradle
          key: ${{ runner.os }}-gradle-${{ hashFiles('android/**/*.gradle*', 'android/gradle/wrapper/gradle-wrapper.properties') }}
          restore-keys: |
            ${{ runner.os }}-gradle-

      - name: Purge stale generated sources
        # Same mitigation as cua-smoke.yml (whose Gradle cache entries this job
        # shares — the cache key/restore-keys are identical): the restore-keys
        # prefix fallback can restore a generated autolinking tree from a
        # previous package id, making compileReleaseJavaWithJavac fail against
        # the old package (ai.opencode.mobile vs cc.agentlabs.opencode). Delete
        # generated sources so prebuild + Gradle regenerate them.
        run: rm -rf android/app/build/generated android/build/generated android/app/build/intermediates

      - name: Install Maestro CLI
        run: |
          curl -Ls "https://get.maestro.mobile.dev" | bash
          echo "$HOME/.maestro/bin" >> $GITHUB_PATH

      - name: Install dependencies & build APK
        env:
          SENTRY_DISABLE_AUTO_UPLOAD: "true"
        run: |
          npm install --legacy-peer-deps
          npx expo prebuild --platform android --no-install
          keytool -genkey -v -keystore android/app/debug.keystore -storepass android -alias androiddebugkey -keypass android -keyalg RSA -keysize 2048 -validity 10000 -dname "CN=Android Debug,O=Android,C=US"
          cd android && ./gradlew assembleRelease

      - name: Start mock opencode servers
        run: |
          set -x
          mkdir -p artifacts/screenshots
          # Normal mode on 4096 (positive flow) and --fail-auth on 4097 (negative flow).
          # 4098 is normal mode + --seed-sessions (two pre-populated sessions in two
          # different directories, for all-sessions.yaml). 4099 is a fresh normal-mode
          # instance shared by directory-picker.yaml and variant-picker.yaml (its fake
          # file tree / provider variants / project list don't affect each other).
          # 4100 is normal mode + --seed-diff (one session with a pre-existing wide
          # edit-diff tool call + wide code block, for diff-scroll.yaml / issue #21).
          # All bind 0.0.0.0 so the emulator can reach them via 10.0.2.2.
          nohup node tests/fixtures/mock-opencode-server.ts --port 4096 > /tmp/mock-4096.log 2>&1 &
          nohup node tests/fixtures/mock-opencode-server.ts --port 4097 --fail-auth > /tmp/mock-4097.log 2>&1 &
          nohup node tests/fixtures/mock-opencode-server.ts --port 4098 --seed-sessions > /tmp/mock-4098.log 2>&1 &
          nohup node tests/fixtures/mock-opencode-server.ts --port 4099 > /tmp/mock-4099.log 2>&1 &
          nohup node tests/fixtures/mock-opencode-server.ts --port 4100 --seed-diff > /tmp/mock-4100.log 2>&1 &
          for port in 4096 4097 4098 4099 4100; do
            for i in $(seq 1 30); do
              if curl -sf --connect-timeout 1 -m 3 "http://127.0.0.1:${port}/global/health" > /dev/null 2>&1; then
                echo "mock server on ${port} responded (may be 401, that's expected on 4097)"
                break
              fi
              # 4097 always 401s -> curl -sf treats that as failure, so also accept "connection made"
              if curl -s --connect-timeout 1 -m 3 -o /dev/null -w '%{http_code}' "http://127.0.0.1:${port}/global/health" 2>/dev/null | grep -qE '^[0-9]+$'; then
                echo "mock server on ${port} is up (got an HTTP response)"
                break
              fi
              if [ "$i" = "30" ]; then
                echo "::error::mock server on port ${port} did not come up in 30s"
                cat "/tmp/mock-${port}.log" || true
                exit 1
              fi
              sleep 1
            done
          done

      - name: Run activation E2E flows on emulator
        uses: reactivecircus/android-emulator-runner@v2
        with:
          api-level: 28
          arch: x86_64
          target: default
          disable-animations: true
          emulator-boot-timeout: 600
          emulator-options: -no-window -no-audio -no-boot-anim -gpu swiftshader_indirect -no-snapshot
          # One script file, not one `sh -c` per line — run-e2e-flows.sh does the
          # adb reverse port-forwarding, runs every flow, and captures logcat on
          # exit. (Running maestro line-by-line here broke `cd` persistence and
          # left no diagnostics.)
          script: bash scripts/run-e2e-flows.sh

      - name: Mock server logs
        if: always()
        run: |
          echo "--- 4096 (normal) ---"; cat /tmp/mock-4096.log || true
          echo "--- 4097 (fail-auth) ---"; cat /tmp/mock-4097.log || true
          echo "--- 4098 (seed-sessions) ---"; cat /tmp/mock-4098.log || true
          echo "--- 4099 (normal, directory-picker + variant-picker) ---"; cat /tmp/mock-4099.log || true
          echo "--- 4100 (seed-diff) ---"; cat /tmp/mock-4100.log || true

      - name: Upload screenshots
        if: always()
        uses: actions/upload-artifact@v7
        with:
          name: activation-e2e-screenshots-${{ github.run_number }}
          path: artifacts/screenshots/*.png
          if-no-files-found: warn

      - name: Upload maestro debug + logcat
        if: always()
        uses: actions/upload-artifact@v7
        with:
          name: activation-e2e-debug-${{ github.run_number }}
          path: |
            artifacts/diag/**
          if-no-files-found: warn
          # Maestro's --debug-output nests the UI-hierarchy dump under a
          # hidden `.maestro/tests/<timestamp>/` directory. upload-artifact
          # excludes dotfiles/dot-directories by default, so every prior run
          # silently uploaded only logcat.txt/probe.txt and dropped the
          # actual hierarchy dumps we need to diagnose flow failures (issue
          # #104) — this was invisible because if-no-files-found: warn
          # doesn't fail the step when SOME files still match.
          include-hidden-files: true