diff --git a/.github/actions/bot-ci-failure/analyze_failure.py b/.github/actions/bot-ci-failure/analyze_failure.py
index dc0e537d..648e5316 100644
--- a/.github/actions/bot-ci-failure/analyze_failure.py
+++ b/.github/actions/bot-ci-failure/analyze_failure.py
@@ -391,6 +391,21 @@ def main():
     ):
         print("::warning::Skipping: No failure logs to analyse.", file=sys.stderr)
         return
+    if not error_log.strip():
+        print("::warning::Skipping: Empty failure logs.", file=sys.stderr)
+        return
+    if not tests_failed and not transient_only:
+        no_failure_patterns = [
+            "No failed jobs found",
+            "Failed jobs found but logs unavailable",
+            "Could not fetch failed jobs",
+        ]
+        if any(p in error_log for p in no_failure_patterns):
+            print(
+                "::notice::No CI failures detected; skipping analysis.",
+                file=sys.stderr,
+            )
+            return
     # Only fetch the full repository code context when automated tests
     # actually failed.  For QA-only or commit-message failures the code
     # is not needed and would waste prompt tokens.
diff --git a/.github/workflows/reusable-bot-ci-failure.yml b/.github/workflows/reusable-bot-ci-failure.yml
index c17dfcc3..d2334b43 100644
--- a/.github/workflows/reusable-bot-ci-failure.yml
+++ b/.github/workflows/reusable-bot-ci-failure.yml
@@ -183,4 +183,8 @@ jobs:
             echo "AI analysis produced no output; skipping comment."
             exit 0
           fi
+          if ! grep -q "##\[error\]" failed_logs.txt 2>/dev/null; then
+            echo "No CI failures detected; skipping comment."
+            exit 0
+          fi
           gh pr comment "$PR_NUM" --repo "$REPO" --body-file solution.md