diff --git a/.github/workflows/test-htcondor.yml b/.github/workflows/test-htcondor.yml
new file mode 100644
index 0000000..eeaedc5
--- /dev/null
+++ b/.github/workflows/test-htcondor.yml
@@ -0,0 +1,110 @@
+name: HTCondor Integration Test
+
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+    branches:
+      - main
+
+jobs:
+  htcondor-test:
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout repository code
+        uses: actions/checkout@v4
+
+      - name: Start HTCondor Mini-Cluster Container
+        run: |
+          # Spin up the container, mapping the local repository workspace to /job
+          docker run -d \
+            --name condor-local \
+            -v "${{ github.workspace }}/sandbox":/job \
+            -w /job \
+            htcondor/mini:el9
+
+      - name: Setup Test User and Directory Permissions
+        run: |
+          # Create the unprivileged user inside the container
+          docker exec condor-local useradd -m test_user
+
+          # Fix ownership so test_user can write stdout/logs back to the host filesystem
+          docker exec condor-local chown -R test_user:test_user /job
+
+      - name: Verify Workspace Inside Container
+        run: |
+          # Diagnostic step to ensure dummy_job.sub and payload.sh are present
+          docker exec condor-local ls -la /job
+
+      - name: Wait for HTCondor Daemons to be Ready
+        run: |
+          echo "Waiting for the HTCondor schedd daemon to boot up..."
+          # Try up to 30 times with a 1-second interval
+          for i in {1..30}; do
+            if docker exec -u test_user condor-local condor_status -schedd > /dev/null 2>&1; then
+              echo "HTCondor is ready!"
+              exit 0
+            fi
+            echo "Still waiting... ($i/30)"
+            sleep 1
+          done
+          echo "Error: HTCondor daemons failed to start within 30 seconds."
+          exit 1
+
+      - name: Submit Dummy Job
+        run: |
+          # Cluster number will always be 1, as we use a fresh container
+          docker exec -u test_user condor-local condor_submit dummy_job.sub
+
+      - name: Wait for Job Completion
+        run: |
+          # Using "condor_watch_q -summary -exit all,done -clusters 1" would have been better,
+          # but it tries to use TTY for rendering information, which fails in GH workflows
+          # So, use a more complex condor_q method...
+          echo "Monitoring Cluster ID: 1..."
+          for i in {1..60}; do
+            # condor_q prints the Cluster ID only if the job is active/idle/running.
+            # If the output is empty, the job has left the queue and is complete.
+            STATUS=$(docker exec -u test_user condor-local condor_q 1 -format "%d" ClusterId)
+
+            if [ -z "$STATUS" ]; then
+              echo "Job left the queue (Completed)!"
+              exit 0
+            fi
+
+            echo "Job is still pending/processing... ($i/60)"
+            sleep 1
+          done
+
+          echo "Error: Job did not finish within 60 seconds."
+          exit 1
+      - name: Verify Job Output Logs
+        run: |
+          # Check the output file written back to the GitHub workspace runner
+          echo "=== Content of job.out ==="
+          cat sandbox/job.out
+
+          # Check the standard error to ensure no underlying bash glitches occurred
+          echo "=== Content of job.err ==="
+          cat sandbox/job.err
+
+          # 1. Assert that our expected string exists in the output file
+          if ! grep -q "Hello from native HTCondor!" sandbox/job.out; then
+            echo "Failure: Expected job output string not found."
+            exit 1
+          fi
+
+          # 2. Assert that the error log is completely empty
+          if [ -s sandbox/job.err ]; then
+            echo "Failure: job.err is not empty! Something went wrong during execution."
+            exit 2
+          fi
+
+          echo "Success: Job completed perfectly with no errors!"
+
+      - name: Cleanup Container
+        if: always() # Ensures the container is torn down even if the tests fail
+        run: |
+          docker rm -f condor-local
diff --git a/sandbox/dummy_job.sub b/sandbox/dummy_job.sub
new file mode 100755
index 0000000..5ed5d1d
--- /dev/null
+++ b/sandbox/dummy_job.sub
@@ -0,0 +1,12 @@
+# dummy_job.sub
+universe                = vanilla
+executable              = payload.sh
+
+output                  = job.out
+error                   = job.err
+log                     = job.log
+
+should_transfer_files   = YES
+when_to_transfer_output = ON_EXIT
+
+queue 1
diff --git a/sandbox/payload.sh b/sandbox/payload.sh
new file mode 100755
index 0000000..9b1be72
--- /dev/null
+++ b/sandbox/payload.sh
@@ -0,0 +1,4 @@
+#!/bin/bash
+echo "Hello from native HTCondor!"
+uname -r
+sleep 5