diff --git a/plugins/code/.claude-plugin/plugin.json b/plugins/code/.claude-plugin/plugin.json
index bc05a88..2e4c6cc 100644
--- a/plugins/code/.claude-plugin/plugin.json
+++ b/plugins/code/.claude-plugin/plugin.json
@@ -1,7 +1,7 @@
 {
   "name": "code",
   "description": "Code and planning framework plugin",
-  "version": "1.14.7",
+  "version": "1.14.8",
   "author": {
     "name": "ClosedLoop",
     "email": "support@closedloop.ai"
diff --git a/plugins/code/scripts/run-loop.sh b/plugins/code/scripts/run-loop.sh
index 1c9f934..5b92ccb 100755
--- a/plugins/code/scripts/run-loop.sh
+++ b/plugins/code/scripts/run-loop.sh
@@ -147,6 +147,11 @@ fail_loop_user_visible() {
 # fail_loop_user_visible.
 detect_spurious_complete() {
   local workdir="$1"
+  # Defaults to the global so the existing single-argument call site is
+  # unchanged; passed explicitly by tests. Non-empty means this run was asked to
+  # draft a plan from a PRD, which is what makes a missing plan.json a broken
+  # promise rather than a run that never owed one.
+  local prd_file="${2-${PRD_FILE:-}}"
   local plan_file="$workdir/plan.json"
   local state_file="$workdir/state.json"
 
@@ -166,6 +171,29 @@ detect_spurious_complete() {
   fi
 
   if [[ ! -f "$plan_file" ]]; then
+    # A --prd run exists to produce plan.json. Claiming COMPLETE without one is
+    # the strongest spurious-completion signal there is, and this branch used to
+    # wave it through: the checks below only validate pendingTasks INSIDE an
+    # existing plan, so "no plan at all" -- the case that actually happens --
+    # was the one case nothing could catch.
+    #
+    # Observed: the orchestrator launched plan-draft-writer in the BACKGROUND,
+    # said "Plan-draft-writer is running in the background. Waiting for
+    # completion.", and that same turn carried the completion promise. The loop
+    # ended, the writer was abandoned mid-flight, post-loop code review passed
+    # vacuously over an empty diff ("the base ref you passed equals HEAD, so
+    # nothing was examined"), and the run exited 0 having produced nothing. The
+    # user got an implementation-plan artifact that looked done and was empty.
+    #
+    # Scoped to runs that were asked for a plan: a run with no PRD never owed
+    # one, and is left alone.
+    if [[ -n "$prd_file" ]]; then
+      jq -n -c \
+        --arg subcode "PLAN_MISSING_AT_COMPLETION" \
+        --arg message "Loop emitted COMPLETE but no plan.json was ever written. The planning phase did not finish -- a background plan-draft-writer that is still running when the completion promise fires is abandoned. Inspect state.json and the loop output, then re-run /code:code to continue." \
+        '{subcode:$subcode,message:$message}'
+      return
+    fi
     echo '{}'
     return
   fi
diff --git a/plugins/code/scripts/tests/test_spurious_complete.sh b/plugins/code/scripts/tests/test_spurious_complete.sh
new file mode 100755
index 0000000..f1ea7f9
--- /dev/null
+++ b/plugins/code/scripts/tests/test_spurious_complete.sh
@@ -0,0 +1,111 @@
+#!/usr/bin/env bash
+# Tests for detect_spurious_complete() in run-loop.sh.
+#
+# The case that motivated these: a PLAN run emitted the completion promise while
+# plan-draft-writer was still running in the BACKGROUND. The loop ended, the
+# writer was abandoned, post-loop code review passed vacuously over an empty
+# diff, and the run exited 0 having written no plan.json at all -- so the user's
+# implementation-plan artifact looked done and was empty. The guard could not
+# see it: its checks only validate pendingTasks INSIDE an existing plan.json.
+#
+# run-loop.sh guards main() with [[ "${BASH_SOURCE[0]}" == "$0" ]], so sourcing
+# it defines the functions without running the loop.
+#
+# Usage:
+#   bash plugins/code/scripts/tests/test_spurious_complete.sh
+#
+# Exit code: 0 if all tests pass, 1 if any test fails.
+
+set -uo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+RUN_LOOP="$SCRIPT_DIR/../run-loop.sh"
+
+PASS_COUNT=0
+FAIL_COUNT=0
+
+pass() {
+  echo "  PASS: $1"
+  PASS_COUNT=$(( PASS_COUNT + 1 ))
+}
+
+fail() {
+  echo "  FAIL: $1 -- $2"
+  FAIL_COUNT=$(( FAIL_COUNT + 1 ))
+}
+
+assert_subcode() {
+  local name="$1" json="$2" expected="$3"
+  local actual
+  actual=$(echo "$json" | jq -r '.subcode // ""' 2>/dev/null || echo "")
+  if [[ "$actual" == "$expected" ]]; then
+    pass "$name (subcode=$actual)"
+  else
+    fail "$name" "expected subcode '$expected', got '$actual' in: $json"
+  fi
+}
+
+assert_not_spurious() {
+  local name="$1" json="$2"
+  local actual
+  actual=$(echo "$json" | jq -r '.subcode // ""' 2>/dev/null || echo "")
+  if [[ -z "$actual" ]]; then
+    pass "$name (not flagged)"
+  else
+    fail "$name" "expected no subcode, got '$actual' in: $json"
+  fi
+}
+
+# shellcheck source=/dev/null
+source "$RUN_LOOP"
+
+echo "detect_spurious_complete"
+
+# --- A --prd run that produced no plan is the reported defect ---------------
+WORKDIR=$(mktemp -d)
+printf '%s' '{"phase":"Phase 1: Planning","status":"IN_PROGRESS"}' > "$WORKDIR/state.json"
+assert_subcode "COMPLETE with no plan.json on a --prd run is spurious" \
+  "$(detect_spurious_complete "$WORKDIR" "/tmp/prd.md")" \
+  "PLAN_MISSING_AT_COMPLETION"
+rm -rf "$WORKDIR"
+
+# --- A run that never owed a plan is left alone ----------------------------
+WORKDIR=$(mktemp -d)
+assert_not_spurious "COMPLETE with no plan.json and no PRD is not spurious" \
+  "$(detect_spurious_complete "$WORKDIR" "")"
+rm -rf "$WORKDIR"
+
+# --- The AWAITING_USER hard stop must keep winning -------------------------
+# A drafted plan parked for review legitimately has no finished plan yet; the
+# new branch must not turn that documented stop into a failure.
+WORKDIR=$(mktemp -d)
+printf '%s' '{"phase":"Phase 1.1","status":"AWAITING_USER"}' > "$WORKDIR/state.json"
+assert_not_spurious "AWAITING_USER outranks the missing-plan check" \
+  "$(detect_spurious_complete "$WORKDIR" "/tmp/prd.md")"
+rm -rf "$WORKDIR"
+
+# --- Existing behaviour: a complete plan is still clean ---------------------
+WORKDIR=$(mktemp -d)
+printf '%s' '{"pendingTasks":[],"openQuestions":[]}' > "$WORKDIR/plan.json"
+assert_not_spurious "a plan with no pending tasks is not spurious" \
+  "$(detect_spurious_complete "$WORKDIR" "/tmp/prd.md")"
+rm -rf "$WORKDIR"
+
+# --- Existing behaviour: pending tasks still flagged, and not as the new code -
+WORKDIR=$(mktemp -d)
+printf '%s' '{"pendingTasks":[{"id":"T-1.1"}],"openQuestions":[]}' > "$WORKDIR/plan.json"
+assert_subcode "pending tasks at completion still flagged" \
+  "$(detect_spurious_complete "$WORKDIR" "/tmp/prd.md")" \
+  "PENDING_TASKS_AT_COMPLETION"
+rm -rf "$WORKDIR"
+
+WORKDIR=$(mktemp -d)
+printf '%s' '{"pendingTasks":[{"id":"T-1.1"}],"openQuestions":[{"id":"Q-1"}]}' > "$WORKDIR/plan.json"
+assert_subcode "pending tasks blocked by open questions still flagged" \
+  "$(detect_spurious_complete "$WORKDIR" "/tmp/prd.md")" \
+  "PENDING_TASKS_BLOCKED_BY_QUESTIONS"
+rm -rf "$WORKDIR"
+
+echo
+echo "passed: $PASS_COUNT  failed: $FAIL_COUNT"
+[[ "$FAIL_COUNT" -eq 0 ]]