Repair two stale shell tests; CI now runs all six

Both tests excluded from CI last commit turned out to be stale rather
than catching regressions. Neither covered script needed changing.

test_music_parallelism.sh built its temp worktree at
scripts/digital_wellbeing/, but music_parallelism.sh resolves its
library as $SCRIPT_DIR/../../lib/common.sh — two levels up, correct
for the real scripts/periodic_background/digital_wellbeing/ location.
From the shallower path that resolved outside the worktree, so the
stubbed common.sh was never sourced and the script fell through to the
real /usr/local/lib/common.sh on the machine. The test was exercising
the live library and ignoring its own MOCK_FOCUS_ACTIVE, which is why
idle mode looked like it waited 15s instead of 30s. The daemon was
correct all along.

test_usage_monitoring_installer_efficiency.sh demanded a literal
'sleep 60'. The logger template no longer polls: it sleeps once until
midnight via seconds_until_next_day, then blocks on wait for the pmon
child — strictly fewer wakeups than the assertion required. Replaced
with assertions on that property so a return to polling still fails.

Allowlist extended from four tests to six.

.github/workflows/shell-tests.yml

@@ -21,16 +21,7 @@ jobs:
       # Only the side-effect-free tests run here. The rest of
       # linux_configuration/tests/ drives pacman, /etc, systemd or adb and
       # cannot run on an Ubuntu runner without root and an Arch container --
-      # see docs note in the accompanying evidence artifact.
-      #
-      # Deliberately NOT in this list (both fail today, and both look like they
-      # are catching real polling-efficiency regressions rather than being
-      # stale -- they need a fix in the scripts they cover, not in the tests):
-      #   test_music_parallelism.sh
-      #       'idle mode uses the idle interval' expects 30s but the daemon
-      #       waits 15s, i.e. it wakes twice as often as intended when idle.
-      #   test_usage_monitoring_installer_efficiency.sh
-      #       'logger template must sleep between day rollover checks'.
+      # see the accompanying evidence artifacts for the full list.
       - name: Run side-effect-free shell tests
         run: |
           set -euo pipefail
@@ -38,7 +29,9 @@ jobs:
             test_android_guardian_post_fs_data.sh
             test_android_guardian_service.sh
             test_i3blocks_persist_common.sh
+            test_music_parallelism.sh
             test_setup_dns_blocker.sh
+            test_usage_monitoring_installer_efficiency.sh
           )
           failed=0
           for t in "${tests[@]}"; do

docs/superpowers/evidence/shell-tests-stale-assertions-2026-07-19.json

@@ -0,0 +1,35 @@
+{
+  "intent": "Repair the last two excluded shell tests so all six side-effect-free tests run in CI, after establishing that both failures were stale tests rather than regressions in the scripts they cover.",
+  "scope": [
+    "linux_configuration/tests/test_music_parallelism.sh: temp-worktree layout",
+    "linux_configuration/tests/test_usage_monitoring_installer_efficiency.sh: one stale assertion replaced with three property assertions",
+    ".github/workflows/shell-tests.yml: allowlist grows from 4 tests to 6",
+    "Non-goal: any change to music_parallelism.sh or install_usage_monitoring.sh — both were found to be correct"
+  ],
+  "changes": [
+    "test_music_parallelism.sh built its temp worktree at scripts/digital_wellbeing/, but the script resolves its library as $SCRIPT_DIR/../../lib/common.sh — two levels up, correct for the real scripts/periodic_background/digital_wellbeing/ location. From the shallower path that resolved outside the worktree, so the stubbed common.sh was never sourced and the script silently fell through to the real /usr/local/lib/common.sh installed on the machine. The test was therefore exercising the live library and ignoring its own MOCK_FOCUS_ACTIVE, which is why idle mode appeared to wait 15s instead of 30s. The worktree now mirrors the real layout.",
+    "test_usage_monitoring_installer_efficiency.sh asserted the logger template contains a literal 'sleep 60'. The template no longer polls: it sleeps once until midnight via seconds_until_next_day and then blocks on wait for the pmon child, which is strictly fewer wakeups. Replaced with three assertions on that property — sleeps until the day boundary, defines seconds_until_next_day, and blocks on wait — so a regression back to polling still fails.",
+    "Workflow allowlist extended to all six side-effect-free tests; the comment about known-red exclusions is removed because none remain."
+  ],
+  "verification": [
+    {
+      "command": "bash linux_configuration/tests/<each of the 6 tests>",
+      "result": "pass",
+      "evidence": "All six PASS, up from four. Before: test_music_parallelism.sh failed 'idle mode uses the idle interval (expected 30, actual 15)'; test_usage_monitoring_installer_efficiency.sh failed 'logger template must sleep between day rollover checks'."
+    },
+    {
+      "command": "grep 'sleep \"$(seconds_until_next_day)\"' install_usage_monitoring.sh",
+      "result": "pass",
+      "evidence": "Confirmed the template sleeps exactly until midnight then blocks on `wait \"$pmon_pid\"`, with no busy-wait anywhere in the loop — better than the sleep-60 polling the old assertion demanded."
+    }
+  ],
+  "risks": [
+    "Both failures were initially reported as polling-efficiency regressions in the covered scripts. That was wrong in both cases, and no script was changed. The lesson is that a test which cannot find its own stub may silently fall back to a real installed library and still appear to run — worth checking first whenever a test's expectations look inexplicably wrong.",
+    "test_music_parallelism.sh now depends on the temp worktree mirroring the real directory depth. Another reorganisation of scripts/ will break it again in the same silent way; the failure will look like a behavioural assertion rather than a path problem.",
+    "The new assertions match on specific template text, so harmless rewording of the logger template would fail them even though behaviour is unchanged."
+  ],
+  "rollback": [
+    "git revert this commit: the two tests return to their stale expectations and CI's allowlist drops back to four tests.",
+    "Validate after rollback: both tests fail again with the original messages, and .github/workflows/shell-tests.yml lists four tests."
+  ]
+}

linux_configuration/tests/test_music_parallelism.sh

@@ -29,8 +29,8 @@ trap cleanup EXIT
 
 WORKTREE="$TMP_DIR/worktree"
 BIN_DIR="$TMP_DIR/bin"
-mkdir -p "$WORKTREE/scripts/digital_wellbeing" "$WORKTREE/scripts/lib" "$BIN_DIR"
-cp "$TARGET_SCRIPT" "$WORKTREE/scripts/digital_wellbeing/music_parallelism.sh"
+mkdir -p "$WORKTREE/scripts/periodic_background/digital_wellbeing" "$WORKTREE/scripts/lib" "$BIN_DIR"
+cp "$TARGET_SCRIPT" "$WORKTREE/scripts/periodic_background/digital_wellbeing/music_parallelism.sh"
 
 cat >"$WORKTREE/scripts/lib/common.sh" <<'EOF'
 #!/bin/bash
@@ -87,7 +87,7 @@ run_case() {
     XDOTOOL_LOG="${XDOTOOL_LOG:-}" \
     PROC_ROOT="$proc_root" \
     MOCK_FOCUS_ACTIVE="$focus_active" \
-    bash "$WORKTREE/scripts/digital_wellbeing/music_parallelism.sh" "$mode" \
+    bash "$WORKTREE/scripts/periodic_background/digital_wellbeing/music_parallelism.sh" "$mode" \
     >/dev/null 2>&1 || true
 
   assert_equals "$expected_wait" "$(<"$wait_log")" "music_parallelism.sh should pick the expected wait interval"

linux_configuration/tests/test_usage_monitoring_installer_efficiency.sh

@@ -26,9 +26,20 @@ printf 'Checking pmon logger template avoids read -t busy-loop pattern...\n'
 ! grep -q 'read -r -t' <<< "$logger_template" \
   || fail 'logger template must not use read -t as sleep surrogate'
 
-printf 'Checking pmon logger template uses sleep-based waiting...\n'
-grep -q 'sleep 60' <<< "$logger_template" \
-  || fail 'logger template must sleep between day rollover checks'
+# The template used to poll with `sleep 60` until the day changed. It now
+# sleeps exactly once, until the next midnight, and blocks on `wait` for the
+# pmon child — strictly fewer wakeups. Assert that property rather than the
+# old literal, so an accidental return to polling still fails the test.
+printf 'Checking pmon logger template sleeps until rollover rather than polling...\n'
+grep -q 'sleep "\$(seconds_until_next_day)"' <<< "$logger_template" \
+  || fail 'logger template must sleep until the next day boundary'
+
+grep -q 'seconds_until_next_day()' <<< "$logger_template" \
+  || fail 'logger template must define seconds_until_next_day'
+
+printf 'Checking pmon logger template blocks on wait instead of spinning...\n'
+grep -qE '^\s*wait "\$pmon_pid"' <<< "$logger_template" \
+  || fail 'logger template must block on wait for the pmon child'
 
 printf 'Checking pmon logger template uses fork-free date builtin...\n'
 grep -q "printf '%(%Y%m%d)T' -1" <<< "$logger_template" \