cleaning up

Author: RoxyFarhad

.github/workflows/agentex-tutorials-test.yml

@@ -20,7 +20,7 @@ jobs:
           all_tutorials=$(find . -name "manifest.yaml" -exec dirname {} \; | sort | sed 's|^\./||')
 
           # Filter out the specified temporal tutorials that are being updated
-          filtered_tutorials=$(echo "$all_tutorials" | grep -v -E "(10_temporal/050_|10_temporal/070_|10_temporal/080_)")
+          filtered_tutorials=$(echo "$all_tutorials" | grep -v -E "(temporal)")
 
           # Convert to JSON array
           tutorials=$(echo "$filtered_tutorials" | jq -R -s -c 'split("\n") | map(select(length > 0))')
@@ -119,40 +119,22 @@ jobs:
           AGENTEX_API_BASE_URL="http://localhost:5003" \
             ./run_agent_test.sh --build-cli "${{ matrix.tutorial }}"
 
-      - name: Upload Test Results
+      - name: Record test result
+        id: test-result
         if: always()
         run: |
-          # Sanitize tutorial name for artifact upload
-          SANITIZED_NAME=$(echo "${{ matrix.tutorial }}" | sed 's/\//-/g')
-          echo "Uploading test results for: ${{ matrix.tutorial }} (as: test-results-$SANITIZED_NAME)"
-
-          # Create a temporary directory with the sanitized name
-          mkdir -p "test-results-$SANITIZED_NAME"
-          cp /tmp/agentex-*.log "test-results-$SANITIZED_NAME/" 2>/dev/null || echo "No log files to copy"
-
-          # Upload using the actions/upload-artifact action
-          echo "artifact-name=test-results-$SANITIZED_NAME" >> $GITHUB_ENV
-
-      - name: Upload Artifact
-        if: always()
-        uses: actions/upload-artifact@v4
-        with:
-          name: ${{ env.artifact-name }}
-          path: test-results-*
-          retention-days: 1
+          if [ "${{ steps.run-test.outcome }}" == "success" ]; then
+            echo "result=passed" >> $GITHUB_OUTPUT
+          else
+            echo "result=failed" >> $GITHUB_OUTPUT
+          fi
 
   test-summary:
     if: always()
     needs: [find-tutorials, test-tutorial]
     runs-on: ubuntu-latest
     name: Test Summary
     steps:
-      - name: Download All Test Results
-        uses: actions/download-artifact@v4
-        with:
-          path: test-results
-          pattern: test-results-*
-
       - name: Generate Test Summary
         run: |
           echo "# 🧪 Tutorial Tests Summary" >> $GITHUB_STEP_SUMMARY
@@ -161,53 +143,52 @@ jobs:
           # Get tutorial list from needs context
           tutorials='${{ needs.find-tutorials.outputs.tutorials }}'
 
-          # Debug: Show what we're working with
-          echo "🔍 DEBUG: Tutorial list from find-tutorials job:"
-          echo "$tutorials"
-          echo ""
-          echo "🔍 DEBUG: Downloaded artifacts:"
-          ls -la test-results/ || echo "No test-results directory found"
-          echo ""
-          echo "🔍 DEBUG: Artifact contents:"
-          find test-results/ -type f -name "*.log" || echo "No log files found"
-          echo ""
+          # Fetch workflow run jobs to get individual test results
+          jobs_json=$(curl -s -H "Authorization: token $GITHUB_TOKEN" \
+            "https://api.github.com/repos/${{ github.repository }}/actions/runs/${{ github.run_id }}/jobs")
 
           # Initialize counters
           total_tutorials=0
           passed_tutorials=0
           failed_tutorials=0
 
-          # Arrays to track results
-          passed_tests=()
-          failed_tests=()
-
           echo "## 📊 Overall Results" >> $GITHUB_STEP_SUMMARY
           echo "" >> $GITHUB_STEP_SUMMARY
 
-          # Process each tutorial result
-          for tutorial_dir in test-results/test-results-*/; do
-            if [ -d "$tutorial_dir" ]; then
-              # Extract sanitized name and convert back to original tutorial path
-              sanitized_name=$(basename "$tutorial_dir" | sed 's/test-results-//')
-              tutorial_name=$(echo "$sanitized_name" | sed 's/-/\//g')
-              total_tutorials=$((total_tutorials + 1))
-
-              # Check if there are any log files in this directory
-              if find "$tutorial_dir" -name "*.log" -type f | grep -q .; then
-                # Determine success/failure based on pytest-specific failure patterns
-                if find "$tutorial_dir" -name "*.log" -exec grep -l "FAILED.*::" {} \; | head -1 >/dev/null || \
-                   find "$tutorial_dir" -name "*.log" -exec grep -l "=== FAILURES ===" {} \; | head -1 >/dev/null || \
-                   find "$tutorial_dir" -name "*.log" -exec grep -l "AssertionError" {} \; | head -1 >/dev/null; then
-                  failed_tutorials=$((failed_tutorials + 1))
-                  failed_tests+=("$tutorial_name")
-                else
-                  passed_tutorials=$((passed_tutorials + 1))
-                  passed_tests+=("$tutorial_name")
-                fi
-              fi
+          # Create temporary files for passed and failed tests
+          > /tmp/passed_tests.txt
+          > /tmp/failed_tests.txt
+
+          # Process each tutorial
+          echo "$tutorials" | jq -r '.[]' | while read -r tutorial_name; do
+            if [ -z "$tutorial_name" ]; then
+              continue
+            fi
+
+            total_tutorials=$((total_tutorials + 1))
+
+            # Find the job for this tutorial
+            job_name="test-$tutorial_name"
+            job_conclusion=$(echo "$jobs_json" | jq -r ".jobs[] | select(.name == \"$job_name\") | .conclusion")
+
+            if [ "$job_conclusion" == "success" ]; then
+              passed_tutorials=$((passed_tutorials + 1))
+              echo "$tutorial_name" >> /tmp/passed_tests.txt
+              echo "✅ $tutorial_name: PASSED"
+            else
+              failed_tutorials=$((failed_tutorials + 1))
+              echo "$tutorial_name" >> /tmp/failed_tests.txt
+              echo "❌ $tutorial_name: FAILED (conclusion: $job_conclusion)"
             fi
           done
 
+          # Read the final counts (since the while loop runs in a subshell)
+          passed_tutorials=$(wc -l < /tmp/passed_tests.txt | tr -d ' ')
+          failed_tutorials=$(wc -l < /tmp/failed_tests.txt | tr -d ' ')
+          total_tutorials=$((passed_tutorials + failed_tutorials))
+
+          echo "Final counts: total=$total_tutorials, passed=$passed_tutorials, failed=$failed_tutorials"
+
           # Show summary stats
           echo "| Status | Count |" >> $GITHUB_STEP_SUMMARY
           echo "|--------|-------|" >> $GITHUB_STEP_SUMMARY
@@ -220,53 +201,31 @@ jobs:
           if [ $passed_tutorials -gt 0 ]; then
             echo "## ✅ Passed Tutorials ($passed_tutorials)" >> $GITHUB_STEP_SUMMARY
             echo "" >> $GITHUB_STEP_SUMMARY
-            for test in "${passed_tests[@]}"; do
-              echo "- ✅ \`$test\`" >> $GITHUB_STEP_SUMMARY
-            done
+            while read -r test; do
+              if [ -n "$test" ]; then
+                echo "- ✅ \`$test\`" >> $GITHUB_STEP_SUMMARY
+              fi
+            done < /tmp/passed_tests.txt
             echo "" >> $GITHUB_STEP_SUMMARY
           fi
 
-          # Show pytest failures only for failed tests
+          # Show failed tests
           if [ $failed_tutorials -gt 0 ]; then
             echo "## ❌ Failed Tutorials ($failed_tutorials)" >> $GITHUB_STEP_SUMMARY
             echo "" >> $GITHUB_STEP_SUMMARY
-            echo '```' >> $GITHUB_STEP_SUMMARY
-
-            # Extract and append pytest failures from each failed test
-            for test in "${failed_tests[@]}"; do
-              # Find the log file for this test (convert back to sanitized name)
-              sanitized_test_name=$(echo "$test" | sed 's/\//-/g')
-              log_file=$(find "test-results/test-results-$sanitized_test_name" -name "*.log" | head -1)
-              if [ -f "$log_file" ]; then
-                echo "================================================================================================" >> $GITHUB_STEP_SUMMARY
-                echo "FAILED: $test" >> $GITHUB_STEP_SUMMARY
-                echo "================================================================================================" >> $GITHUB_STEP_SUMMARY
-
-                # Extract pytest output between the delimiters, or show pytest summary if no delimiters
-                if grep -q "========== PYTEST OUTPUT ==========" "$log_file"; then
-                  sed -n '/========== PYTEST OUTPUT ==========/,/========== END PYTEST OUTPUT ==========/p' "$log_file" | \
-                    sed '1d;$d' >> $GITHUB_STEP_SUMMARY
-                else
-                  # If no delimiters, try to extract pytest-related lines
-                  grep -E "(FAILED|ERROR|AssertionError|collected.*items|=====.*=====|::.*FAILED)" "$log_file" >> $GITHUB_STEP_SUMMARY || \
-                    echo "No pytest output found in log file" >> $GITHUB_STEP_SUMMARY
-                fi
-                echo "" >> $GITHUB_STEP_SUMMARY
-              else
-                echo "================================================================================================" >> $GITHUB_STEP_SUMMARY
-                echo "FAILED: $test (No log file found)" >> $GITHUB_STEP_SUMMARY
-                echo "================================================================================================" >> $GITHUB_STEP_SUMMARY
-                echo "" >> $GITHUB_STEP_SUMMARY
+            echo "Check the individual job logs for failure details." >> $GITHUB_STEP_SUMMARY
+            echo "" >> $GITHUB_STEP_SUMMARY
+            while read -r test; do
+              if [ -n "$test" ]; then
+                echo "- ❌ \`$test\`" >> $GITHUB_STEP_SUMMARY
               fi
-            done
-
-            echo '```' >> $GITHUB_STEP_SUMMARY
+            done < /tmp/failed_tests.txt
             echo "" >> $GITHUB_STEP_SUMMARY
           fi
 
           # Set exit code based on results
           if [ $failed_tutorials -gt 0 ]; then
-            echo "❌ Some tutorials failed. Check the details above." >> $GITHUB_STEP_SUMMARY
+            echo "❌ Some tutorials failed. Check the individual job logs for details." >> $GITHUB_STEP_SUMMARY
             exit 1
           else
             echo "🎉 All tutorials passed successfully!" >> $GITHUB_STEP_SUMMARY

examples/tutorials/10_agentic/00_base/040_other_sdks/tests/test_agent.py

@@ -114,7 +114,7 @@ async def test_send_event_and_poll_simple_query(self, client: AsyncAgentex, agen
                 break
 
         # Verify state has been updated by polling the states for 10 seconds
-        for i in range(10):
+        for i in range(20):
             if i == 9:
                 raise Exception("Timeout waiting for state updates")
             states = await client.states.list(agent_id=agent_id, task_id=task.id)
@@ -187,7 +187,12 @@ async def test_multi_turn_conversation_with_state(self, client: AsyncAgentex, ag
             sleep_interval=1.0,
         ):
             assert isinstance(message, TaskMessage)
-            if message.content and message.content.type == "text" and message.content.author == "agent" and message.content.content:
+            if (
+                message.content
+                and message.content.type == "text"
+                and message.content.author == "agent"
+                and message.content.content
+            ):
                 break
 
         ## keep polling the states for 10 seconds for the input_list and turn_number to be updated
@@ -216,7 +221,12 @@ async def test_multi_turn_conversation_with_state(self, client: AsyncAgentex, ag
             timeout=30,
             sleep_interval=1.0,
         ):
-            if message.content and message.content.type == "text" and message.content.author == "agent" and message.content.content:
+            if (
+                message.content
+                and message.content.type == "text"
+                and message.content.author == "agent"
+                and message.content.content
+            ):
                 response_text = message.content.content.lower()
                 assert "blue" in response_text
                 found_response = True
@@ -273,7 +283,10 @@ async def stream_messages() -> None:
                 # For full messages, content is at the top level
                 # For delta messages, we need to check parent_task_message
                 if msg_type == "full":
-                    if event.get("content", {}).get("type") == "text" and event.get("content", {}).get("author") == "user":
+                    if (
+                        event.get("content", {}).get("type") == "text"
+                        and event.get("content", {}).get("author") == "user"
+                    ):
                         user_message_found = True
                 elif msg_type == "done":
                     break

examples/tutorials/run_agent_test.sh

@@ -263,7 +263,7 @@ run_test() {
 
 
     # Run the tests with retry mechanism
-    local max_retries=3
+    local max_retries=5
     local retry_count=0
     local exit_code=1