diff --git a/tests/run-godot b/tests/run-godot
index d37a1c45f..0ef7cb8cc 100755
--- a/tests/run-godot
+++ b/tests/run-godot
@@ -3,7 +3,8 @@
 #
 # Exit: 0 = all pass, non-zero = failure.
 # Stdout: JSON summary on ONE line, plus a pointer to the full log.
-#   {"suite":"godot","total":N,"passed":N,"failed":N,"duration_ms":N,"log":"/tmp/sr-run-godot.log"}
+#   {"suite":"godot","total":N,"passed":N,"failed":N,"skipped":N,"duration_ms":N,"log":"..."}
+#   (on a harness failure: same JSON + "harness_error":"load_error"|"no_tests"; exit 2)
 #   (on timeout: same JSON + "timeout":true, "timeout_sec":N; exit code 124)
 # Stderr: a short hint line pointing at the log. The Godot/gdUnit4 output
 #   does NOT stream to stdout or stderr — it is captured to the log file.
@@ -85,20 +86,47 @@ if [[ "$EXIT_CODE" -eq 124 || "$EXIT_CODE" -eq 137 ]]; then
     TIMED_OUT=true
 fi
 
-# gdUnit4 outputs per-suite statistics: "N test cases | X errors | Y failures | ..."
-TOTAL=0; PASSED=0; FAILED=0
-STATS_LINES=$(grep -oE "[0-9]+ test cases \| [0-9]+ errors \| [0-9]+ failures" "$LOG_FILE" || true)
-if [[ -n "$STATS_LINES" ]]; then
-    TOTAL=$(echo "$STATS_LINES" | grep -oE '^[0-9]+' | awk '{s+=$1} END {print s}')
-    ERRORS=$(echo "$STATS_LINES" | grep -oE '[0-9]+ errors' | grep -oE '^[0-9]+' | awk '{s+=$1} END {print s}')
-    FAILURES=$(echo "$STATS_LINES" | grep -oE '[0-9]+ failures' | grep -oE '^[0-9]+' | awk '{s+=$1} END {print s}')
-    FAILED=$(( ${ERRORS:-0} + ${FAILURES:-0} ))
-    PASSED=$(( TOTAL - FAILED ))
+# gdUnit4 prints one "Statistics:" line PER SUITE and then a single
+# "Overall Summary:" line whose numbers are the sum of all of them.
+#
+# BUG FIXED 2026-07-27: the old pattern matched BOTH shapes and summed all of
+# them, so every reported total was exactly DOUBLE — the real number plus the
+# summary that already contained it. A full run reported 3,660 tests against an
+# actual 1,830, and a 26-test suite reported 52. It had been wrong for as long
+# as the summary line has existed, and it was invisible precisely because it
+# doubled uniformly: nothing ever looked inconsistent, only large.
+#
+# Prefer the Overall Summary — it is gdUnit4's own arithmetic over the whole
+# run, so it cannot disagree with itself. Fall back to summing per-suite lines
+# only if it is absent (older gdUnit4, or a run that died mid-way).
+# gdUnit4 colourises its output, and the escape sequences sit BETWEEN the
+# fields of the summary line — so every pattern here must run against a
+# de-ANSI'd copy. Matching the raw log silently falls through to the weaker
+# "Executed test cases" fallback, which cannot see skips and reported a fully
+# skipped suite as 26 FAILED (caught 2026-07-27, immediately after the
+# double-count fix — same log, second lie).
+CLEAN_LOG="${LOG_FILE%.log}.clean.log"
+sed 's/\x1b\[[0-9;]*m//g' "$LOG_FILE" > "$CLEAN_LOG" 2>/dev/null || cp "$LOG_FILE" "$CLEAN_LOG"
+
+TOTAL=0; PASSED=0; FAILED=0; SKIPPED=0
+SUMMARY_LINE=$(grep -oE "Overall Summary: [0-9]+ test cases \| [0-9]+ errors \| [0-9]+ failures \| [0-9]+ flaky \| [0-9]+ skipped" "$CLEAN_LOG" | tail -1 || true)
+if [[ -z "$SUMMARY_LINE" ]]; then
+    SUMMARY_LINE=$(grep -oE "Statistics: [0-9]+ test cases \| [0-9]+ errors \| [0-9]+ failures \| [0-9]+ flaky \| [0-9]+ skipped" "$CLEAN_LOG" || true)
+fi
+if [[ -n "$SUMMARY_LINE" ]]; then
+    _sum() { echo "$SUMMARY_LINE" | grep -oE "[0-9]+ $1" | grep -oE '^[0-9]+' | awk '{s+=$1} END {print s+0}'; }
+    TOTAL=$(echo "$SUMMARY_LINE" | grep -oE '[0-9]+ test cases' | grep -oE '^[0-9]+' | awk '{s+=$1} END {print s+0}')
+    FAILED=$(( $(_sum errors) + $(_sum failures) ))
+    SKIPPED=$(_sum skipped)
+    # A skipped test is NOT a passing test. Counting it as one is the same
+    # false-green shape the guards below exist to prevent — and it matters now
+    # that whole suites are deliberately skipped (test_character_visual_sprint28).
+    PASSED=$(( TOTAL - FAILED - SKIPPED ))
 fi
 
 # Fallback: parse "Executed test cases : (X/N)" for total if stats parse failed
 if [[ "$TOTAL" -eq 0 ]]; then
-    EXEC_LINE=$(grep -oE "Executed test cases : \([0-9]+/[0-9]+\)" "$LOG_FILE" | tail -1 || true)
+    EXEC_LINE=$(grep -oE "Executed test cases : \([0-9]+/[0-9]+\)" "$CLEAN_LOG" | tail -1 || true)
     if [[ -n "$EXEC_LINE" ]]; then
         TOTAL=$(echo "$EXEC_LINE" | grep -oE '/[0-9]+\)' | grep -oE '[0-9]+')
         PASSED=$(echo "$EXEC_LINE" | grep -oE '\([0-9]+/' | grep -oE '[0-9]+')
@@ -116,8 +144,8 @@ fi
 # Neither is a test RESULT, so neither may be reported as one. A run that
 # executed zero tests is never a pass, and a run that could not load a suite
 # is a harness failure regardless of how many other suites went green.
-LOAD_ERROR_COUNT=$(grep -c 'Failed to load script' "$LOG_FILE" 2>/dev/null || true)
-LOAD_ERROR_LIST=$(grep -oE 'Failed to load script "[^"]+"' "$LOG_FILE" 2>/dev/null | sort -u || true)
+LOAD_ERROR_COUNT=$(grep -c 'Failed to load script' "$CLEAN_LOG" 2>/dev/null || true)
+LOAD_ERROR_LIST=$(grep -oE 'Failed to load script "[^"]+"' "$CLEAN_LOG" 2>/dev/null | sort -u || true)
 HARNESS_ERROR=""
 if [[ "${LOAD_ERROR_COUNT:-0}" -gt 0 ]]; then
     HARNESS_ERROR="load_error"
@@ -135,8 +163,8 @@ elif [[ -n "$HARNESS_ERROR" ]]; then
     printf '{"suite":"godot","total":%d,"passed":%d,"failed":%d,"duration_ms":%d,"harness_error":"%s","log":"%s"}\n' \
         "${TOTAL:-0}" "${PASSED:-0}" "${FAILED:-0}" "$DURATION_MS" "$HARNESS_ERROR" "$LOG_FILE"
 else
-    printf '{"suite":"godot","total":%d,"passed":%d,"failed":%d,"duration_ms":%d,"log":"%s"}\n' \
-        "${TOTAL:-0}" "${PASSED:-0}" "${FAILED:-0}" "$DURATION_MS" "$LOG_FILE"
+    printf '{"suite":"godot","total":%d,"passed":%d,"failed":%d,"skipped":%d,"duration_ms":%d,"log":"%s"}\n' \
+        "${TOTAL:-0}" "${PASSED:-0}" "${FAILED:-0}" "${SKIPPED:-0}" "$DURATION_MS" "$LOG_FILE"
 fi
 
 # Hint line on stderr. Stays short. LLM callers: read this literally.