feat: live app log console, cross-filter, JSONL export

Backend:
- /ws/app-logs — real-time app/cot log stream (raw JSONL)
- /api/logs/recent — REST snapshot of ring buffer
- GET /tasks/{id}/logs/export — streaming JSONL export with CoT parse + redaction
- Thread-safe ring buffer (seq-based polling) replaces unsafe asyncio.Queue
- GIL-friendly multi-row Core insert for log persistence
- task_id ContextVar propagates into CotJsonFormatter for CoT correlation
- Hot-apply logging level on settings update (FR-005)
- Buffer trim under DEBUG floods; drop DEBUG first, preserve ERROR/WARNING
- List projection (include_result=False) keeps reports list slim
- Security event consolidated to single REASON atom

Frontend:
- ReportsLogModel + ReportsLogPanel — full live JSONL console
- Cross-filter pinning: Tasks → Logs tab with task chip badges
- LogEntryRow — CoT-aware rendering (marker icons, expandable payload)
- TaskFilterChip, taskChipMeta — scannable type/id/env chips
- Global drawer push via CSS variable (lg+ padding, not overlay)
- i18n en/ru for all log console strings
- Ctrl+A in log panel selects only log lines (window-level handler)

QA fixes:
- Svelte 5 reactivity: SvelteDate/SvelteSet/SvelteURLSearchParams
- Fix seed_trace_id shadowing (F823) in lifecycle.py
- Remove dead code (selectedEnvironment, goToReportsPage)
- Add @BRIEF to C2 test functions, missing {#each} keys
- Remove unused import json as _json from app.py

All 3200+ frontend tests pass; backend lint clean.

backend/src/services/reports/normalizer.py

@@ -108,6 +108,8 @@ def extract_error_context(task: Task, report_status: ReportStatus) -> ErrorConte
 # @BRIEF Convert one Task to canonical TaskReport envelope.
 # @PRE  task has valid id and plugin_id fields.
 # @POST  Returns TaskReport with required fields and deterministic fallback behavior.
+# @POST  When include_result=False (list projection), omits details.result so list
+#        payloads stay small; detail endpoint keeps include_result=True.
 #
 # @TEST_CONTRACT  NormalizeTaskReport ->
 # {
@@ -115,14 +117,15 @@ def extract_error_context(task: Task, report_status: ReportStatus) -> ErrorConte
 #   invariants: [
 #       "Returns a valid TaskReport object",
 #       "Maps TaskStatus to ReportStatus deterministically",
-#       "Extracts ErrorContext for FAILED/PARTIAL tasks"
+#       "Extracts ErrorContext for FAILED/PARTIAL tasks",
+#       "List projection (include_result=False) never embeds task.result"
 #   ]
 # }
 # @TEST_FIXTURE  valid_task -> {"task": "MockTask(id='1', plugin_id='superset-migration', status=TaskStatus.SUCCESS)"}
 # @TEST_EDGE  task_with_error -> {"task": "MockTask(status=TaskStatus.FAILED, logs=[LogEntry(level='ERROR', message='Failed')])"}
 # @TEST_EDGE  unknown_plugin_type -> {"task": "MockTask(plugin_id='unknown-plugin', status=TaskStatus.PENDING)"}
 # @TEST_INVARIANT  deterministic_normalization -> verifies: [valid_task, task_with_error, unknown_plugin_type]
-def normalize_task_report(task: Task) -> TaskReport:
+def normalize_task_report(task: Task, *, include_result: bool = True) -> TaskReport:
     with belief_scope("normalize_task_report"):
         task_type = resolve_task_type(task.plugin_id)
         report_status = status_to_report_status(task.status)
@@ -146,6 +149,8 @@ def normalize_task_report(task: Task) -> TaskReport:
         if not updated_at:
             updated_at = started_at or datetime.now(UTC)
 
+        # List projection: keep only lightweight profile metadata (~150 B).
+        # Full task.result (LLM logs, backup dashboards, …) belongs on detail only.
         details: dict[str, Any] = {
             "profile": {
                 "display_label": profile.get("display_label"),
@@ -153,8 +158,11 @@ def normalize_task_report(task: Task) -> TaskReport:
                 "icon_token": profile.get("icon_token"),
                 "emphasis_rules": profile.get("emphasis_rules", []),
             },
-            "result": task.result if task.result is not None else {"note": "Not provided"},
         }
+        if include_result:
+            details["result"] = (
+                task.result if task.result is not None else {"note": "Not provided"}
+            )
 
         source_ref: dict[str, Any] = {}
         if isinstance(task.params, dict):