feat(newos4): R5 stale-artefact rule + dashboard queues a decision once

Feedback queue triage 2026-11-02: 23 reports, 22 decided, 1 open -- and the open
one (RPT-20261029-115241, "[Terminal] Error in ShellModule.test: {\"silent\":true}")
was already fixed in v4.6.135 (4472c81, modules/core/terminal.js:1966) yet could
never be cleared by automation.

Root cause: identifiers() keeps only generic class names (Terminal, ShellModule),
both of which necessarily stay in the code, so R2 (every identifier absent) cannot
fire; R4 needs a stack and this report has none. A stack-less
"Class.method: {json}" console.error whose emitter gets *guarded* -- rather than
deleted -- therefore sits "awaiting decision" forever.

- newos4-feedback-autoreview.py: new R5 stale artefact -- auto report, no stack,
  every error count <= 1, and nothing newer than STALE_DAYS=5 in
  data/debug/errors.json -> resolve with the measured quiet_days as evidence.
  New --errors <path> knob so the rule can be run against fixtures.
  RED (pristine copy): "nothing provable (1 left for the operator)"
  GREEN (aged fixture): "resolve ... quiet 5.3d -- reopen if it recurs"
  NEG (live store 4.3d, and a report with count 21): both still refuse.
- admin/feedback.html: markDecided() patches the local snapshot, so a queued
  decision leaves the default "Awaiting only" list immediately instead of after
  the 10-minute sync rebuild. Measured cause of 3x approve on one plan in 3 min
  (10-30 16:03/16:05/16:05) and 6 items with duplicate rows.
- AGENTS.md: documents R5 and the fixture knob (also carries the already-live
  2026-11-02 cookie-notice paragraph that was pending in the tree).

1 files changed, 47 insertions, 1 deletions

diff --git a/scripts/newos4-feedback-autoreview.py b/scripts/newos4-feedback-autoreview.py
index e99a516..2021de3 100755
--- a/scripts/newos4-feedback-autoreview.py
+++ b/scripts/newos4-feedback-autoreview.py
@@ -17,6 +17,8 @@ MODULES = f'{ROOT}/spaces/newos4/modules'
 DECISIONS = f'{ROOT}/spaces/newos4/data/feedback/decisions.jsonl'
 AUTONOMY = f'{ROOT}/spaces/newos4/data/feedback/autonomy.json'
 SUMMARY = f'{ROOT}/spaces/newos4/data/feedback/autoreview-latest.json'
+ERRORS = f'{ROOT}/spaces/newos4/data/debug/errors.json'   # symptom store (count/last_seen)
+STALE_DAYS = 5                                            # quiet this long => stale artefact
 CAPS = {'safe': 2, 'aggressive': 5}
 
 NOISE = re.compile(r'fetch failed|failed to fetch|network ?error|load failed|user aborted'
@@ -115,6 +117,32 @@ def token_history(tokens):
     return out
 
 
+def _age_days(ts):
+    """Days since an ISO-8601 stamp; tolerant of Z, fractional seconds and bare datetimes."""
+    s = str(ts or '').strip().replace('Z', '+00:00')
+    if not s:
+        return None
+    try:
+        dt = datetime.fromisoformat(s)
+    except ValueError:
+        return None
+    if dt.tzinfo is None:
+        dt = dt.replace(tzinfo=timezone.utc)
+    return (datetime.now(timezone.utc) - dt).total_seconds() / 86400.0
+
+
+def error_stats(msg):
+    """count/last_seen for one exact message in the debug error store; None when unknown."""
+    rows = load_json(ERRORS, [])
+    if not isinstance(rows, list):
+        return None
+    key = str(msg or '').strip()
+    for e in rows:
+        if isinstance(e, dict) and str(e.get('message') or '').strip() == key:
+            return {'count': int(e.get('count') or 0), 'last_seen': str(e.get('last_seen') or '')}
+    return None
+
+
 def fingerprints():
     """fingerprint -> id, for reports that already exist (duplicate detection)."""
     out = {}
@@ -175,12 +203,28 @@ def classify(rep, dec, fps, cap_left):
     if stack and present and cap_left <= 0:
         return None, 'daily auto-approve cap reached — left for the operator', {'present_tokens': present}
 
+    # R5 stale artefact: one occurrence that never came back, and no stack to act on.
+    # Distinct from R2 (identifier removed) — here the emitter is still there but the symptom is not.
+    if errs and not stack and all(int(e.get('count') or 0) <= 1 for e in errs):
+        stats = [error_stats(str(e.get('message') or '')) for e in errs]
+        stats = [s for s in stats if s and s.get('last_seen')]
+        ages = [_age_days(s['last_seen']) for s in stats]
+        if len(stats) == len(errs) and ages and all(a is not None for a in ages):
+            quiet = max(a for a in ages if a is not None)
+            if quiet >= STALE_DAYS and all(s['count'] <= 1 for s in stats):
+                return 'resolve', (f'stale single-occurrence auto report: no stack, count 1, '
+                                   f'quiet {quiet:.1f}d — reopen if it recurs'), {
+                    'counts': [s['count'] for s in stats],
+                    'last_seen': [s['last_seen'] for s in stats],
+                    'quiet_days': round(quiet, 1),
+                    'emitter_tokens': tokens}
+
     return None, 'not provable by the rules — left for the operator', {
         'tokens': tokens, 'present': present, 'stack': stack}
 
 
 def main():
-    global REPORTS, MODULES, DECISIONS, ROOTDIR
+    global REPORTS, MODULES, DECISIONS, ROOTDIR, ERRORS
     ap = argparse.ArgumentParser()
     ap.add_argument('--root', default=ROOT)
     ap.add_argument('--dry-run', action='store_true')
@@ -188,10 +232,12 @@ def main():
     ap.add_argument('--decisions', default=DECISIONS)
     ap.add_argument('--modules', default=MODULES)
     ap.add_argument('--cap', type=int, default=None)
+    ap.add_argument('--errors', default=ERRORS, help='debug error store (fixtures)')
     ap.add_argument('--include-plans', action='store_true', help='never used in production')
     a = ap.parse_args()
     ROOTDIR = a.root
     REPORTS, MODULES, DECISIONS = a.reports_dir, a.modules, a.decisions
+    ERRORS = a.errors
 
     m, cap = mode(a.cap)
     if m == 'off':