Add E2E test pipeline for azure.ai.agents extension

MCP-tool-driven workflow for running cli-interactive-tester scenarios.
Manual dispatch only (workflow_dispatch) with tier selection (0, 0+1, 0+1+2).

Key design:
- All scenarios executed via cli-interactive-tester MCP tool (not shell parsing)
- Tool installed via git clone + pip install -e from coreai-microsoft repo
- Checkout hardcoded to trangevi/test-scenarios (until PR Azure#8524 merges)
- ubuntu-22.04 runner (consistent with existing pipelines)
- profile.local.yaml generated from GitHub secrets at runtime
- Tier 2 includes always-run teardown step for resource cleanup
- Results uploaded as artifacts

Blocking: python -m auto_test_tool.runner batch mode needs to be confirmed
or implemented. Without it, scenarios cannot run headlessly in CI.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: Jian Wu

.github/workflows/e2e-ext-azure-ai-agents.yml

@@ -0,0 +1,140 @@
+name: "ext-azure-ai-agents-e2e"
+
+# E2E tests for the azure.ai.agents CLI extension.
+# Driven by cli-interactive-tester MCP tool (tmux-based scenario execution).
+#
+# Manual trigger only — switch to PR-trigger once pipeline is stable.
+# Scenarios live on branch trangevi/test-scenarios until PR #8524 merges.
+
+on:
+  workflow_dispatch:
+    inputs:
+      tier:
+        description: "Which tiers to run"
+        type: choice
+        options:
+          - "0"
+          - "0+1"
+          - "0+1+2"
+        default: "0"
+      confirm_tier2_cost:
+        description: "Confirm Tier 2 Azure costs (~$2-5)"
+        type: boolean
+        default: false
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.run_id }}
+  cancel-in-progress: true
+
+permissions:
+  contents: read
+
+env:
+  SCENARIOS_DIR: cli/azd/extensions/azure.ai.agents/tests/cli-interactive-tester-scenarios
+  AZD_AGENTS_FIXTURES: cli/azd/extensions/azure.ai.agents/tests/cli-interactive-tester-scenarios/fixtures
+  # TODO: change to 'main' after PR #8524 merges
+  TARGET_BRANCH: trangevi/test-scenarios
+
+jobs:
+  e2e-test:
+    name: "E2E scenarios (Tier ${{ inputs.tier }})"
+    runs-on: ubuntu-22.04
+    timeout-minutes: 90
+    steps:
+      - uses: actions/checkout@v6
+        with:
+          # TODO: remove ref after PR #8524 merges to main
+          ref: ${{ env.TARGET_BRANCH }}
+
+      - uses: actions/setup-go@v6
+        with:
+          go-version-file: "cli/azd/go.mod"
+
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+
+      - name: Build azd + install extension
+        working-directory: cli/azd
+        run: |
+          go build -o ./azd .
+          export PATH="$PWD:$PATH"
+          azd extension install azure.ai.agents --source ./extensions/azure.ai.agents
+
+      - name: Install cli-interactive-tester
+        run: |
+          sudo apt-get install -y tmux
+          git clone https://github.com/coreai-microsoft/cli-interactive-tester.git /tmp/cli-interactive-tester
+          cd /tmp/cli-interactive-tester
+          pip install -e .
+          python -c "from auto_test_tool.mcp_server import main; print('MCP tool ready')"
+
+      - name: Install uv (for Tier 2 run-local scenario)
+        if: contains(inputs.tier, '2')
+        uses: astral-sh/setup-uv@v6
+
+      - name: Azure Login
+        uses: azure/login@v2
+        with:
+          client-id: ${{ secrets.AZURE_CLIENT_ID }}
+          tenant-id: ${{ secrets.AZURE_TENANT_ID }}
+          subscription-id: ${{ secrets.AZURE_SUBSCRIPTION_ID }}
+
+      - name: GitHub CLI Auth
+        run: echo "${{ secrets.GH_TOKEN }}" | gh auth login --with-token
+
+      - name: Create test profile
+        working-directory: ${{ env.SCENARIOS_DIR }}
+        run: |
+          cat > profile.local.yaml <<EOF
+          prefix: "ci-${{ github.run_number }}"
+          subscription: "${{ secrets.AZURE_SUBSCRIPTION_ID }}"
+          region: "North Central US"
+          foundry_project_endpoint: "${{ secrets.FOUNDRY_PROJECT_ENDPOINT }}"
+          EOF
+
+      - name: Run scenarios via MCP tool
+        working-directory: ${{ env.SCENARIOS_DIR }}
+        env:
+          AZD_AGENTS_FIXTURES: ${{ github.workspace }}/${{ env.AZD_AGENTS_FIXTURES }}
+          UV_HTTP_TIMEOUT: "300"
+        run: |
+          export PATH="${{ github.workspace }}/cli/azd:$PATH"
+          # Drive all scenarios through cli-interactive-tester.
+          # Each scenario is: load_scenario → run_pre_hooks → start_session
+          #   → [driver accomplishes goals] → finish_session → run_post_hooks
+          #
+          # TODO: confirm batch runner CLI exists. If not, need to implement
+          # a wrapper that calls auto_test_tool internal APIs directly.
+          python -m auto_test_tool.runner \
+            --scenarios-dir . \
+            --profile profile.yaml \
+            --profile-local profile.local.yaml \
+            --tier ${{ inputs.tier }} \
+            --output results.json
+
+      - name: Upload results
+        if: always()
+        uses: actions/upload-artifact@v4
+        with:
+          name: e2e-results-tier-${{ inputs.tier }}
+          path: |
+            ${{ env.SCENARIOS_DIR }}/results.json
+            ${{ env.SCENARIOS_DIR }}/reports/
+          retention-days: 30
+
+      - name: Teardown (Tier 2 always cleanup)
+        if: always() && contains(inputs.tier, '2')
+        working-directory: ${{ env.SCENARIOS_DIR }}
+        env:
+          AZD_AGENTS_FIXTURES: ${{ github.workspace }}/${{ env.AZD_AGENTS_FIXTURES }}
+        run: |
+          export PATH="${{ github.workspace }}/cli/azd:$PATH"
+          # Ensure Azure resources are cleaned up even on failure
+          python -m auto_test_tool.runner \
+            --scenarios-dir . \
+            --scenario 2Z-teardown-down.yaml \
+            --profile profile.yaml \
+            --profile-local profile.local.yaml || true
+
+