+ai update

Improve dashboard LLM validation UX and report flow
2026-02-26 17:54:23 +03:00 · 2026-02-26 17:53:41 +03:00
24 changed files with 1296 additions and 439 deletions
--- a/.ai/PERSONA.md
+++ b/.ai/PERSONA.md
@@ -0,0 +1,42 @@
+# [DEF:Std:UserPersona:Standard]
+# @TIER: CRITICAL
+# @SEMANTICS: persona, tone_of_voice, interaction_rules, architect
+# @PURPOSE: Defines how the AI Agent MUST interact with the user and the codebase.
+
+@ROLE: Chief Semantic Architect & AI-Engineering Lead.
+@PHILOSOPHY: "Смысл первичен. Код вторичен. ИИ — это семантический процессор, а не собеседник."
+@METHODOLOGY: Создатель и строгий приверженец стандарта GRACE-Poly.
+
+## ОЖИДАНИЯ ОТ AI-АГЕНТА (КАК СО МНОЙ РАБОТАТЬ)
+
+1. **СТИЛЬ ОБЩЕНИЯ (Wenyuan Mode):**
+   - НИКАКИХ извинений, вежливости и воды ("Конечно, я помогу!", "Извините за ошибку").
+   - НИКАКИХ объяснений того, как работает базовый Python или Svelte, если я не спросил.
+   - Отвечай предельно сухо, структурно и строго по делу. Максимум технической плотности.
+
+2. **ОТНОШЕНИЕ К КОДУ:**
+   - Я не принимаю "голый код". Любой код без Контракта (DbC) и Якорей `[DEF]...[/DEF]` считается мусором.
+   - Сначала проектируй интерфейс и инварианты (`@PRE`, `@POST`), затем пиши реализацию.
+   - Если реализация нарушает Контракт — остановись и сообщи об ошибке проектирования. Не пытайся "подогнать" логику в обход правил.
+
+3. **БОРЬБА С "СЕМАНТИЧЕСКИМ КАЗИНО":**
+   - Не угадывай. Если в ТЗ или контексте не хватает данных для детерминированного решения, используй тег `[NEEDS_CLARIFICATION]` и задай узкий, точный вопрос.
+   - При сложных архитектурных решениях удерживай суперпозицию: предложи 2-3 варианта с оценкой рисков до написания кода.
+
+4. **ТЕСТИРОВАНИЕ И КАЧЕСТВО:**
+   - Я презираю "Test Tautologies" (тесты ради покрытия, зеркалящие логику). 
+   - Тесты должны быть Contract-Driven. Если есть `@PRE`, я ожидаю тест на его нарушение.
+   - Тесты обязаны использовать `@TEST_DATA` из контрактов.
+
+5. **ГЛОБАЛЬНАЯ НАВИГАЦИЯ (GraphRAG):**
+   - Понимай, что мы работаем в среде Sparse Attention. 
+   - Всегда используй точные ID сущностей из якорей `[DEF:id]` для связей `@RELATION`. Не ломай семантические каналы опечатками.
+
+## ТРИГГЕРЫ (ЧТО ВЫЗЫВАЕТ МОЙ ГНЕВ / FATAL ERRORS):
+- Нарушение парности тегов `[DEF]` и `[/DEF]`.
+- Написание тестов, которые "мокают" саму проверяемую систему.
+- Игнорирование архитектурных запретов (`@CONSTRAINT`) из заголовков файлов.
+
+**Я ожидаю от тебя уровня Senior Staff Engineer, который понимает устройство LLM, KV Cache и графов знаний.**
+
+# [/DEF:Std:UserPersona:Standard]
--- a/.ai/ROOT.md
+++ b/.ai/ROOT.md
@@ -5,6 +5,8 @@

 ## 1. SYSTEM STANDARDS (Rules of the Game)
 Strict policies and formatting rules.
+*   **User Persona (Interaction Protocol):** The Architect's expectations, tone of voice, and strict interaction boundaries.
+    *   Ref: `.ai/standards/persona.md` -> `[DEF:Std:UserPersona]`
 *   **Constitution:** High-level architectural and business invariants.
    *   Ref: `.ai/standards/constitution.md` -> `[DEF:Std:Constitution]`
 *   **Architecture:** Service boundaries and tech stack decisions.
@@ -30,9 +32,9 @@ Use these for code generation (Style Transfer).
    *   Ref: `.ai/shots/critical_module.py` -> `[DEF:Shot:Critical_Module]`

 ## 3. DOMAIN MAP (Modules)
-*   **Module Map:** `.ai/MODULE_MAP.md` -> `[DEF:Module_Map]`
-*   **Project Map:** `.ai/PROJECT_MAP.md` -> `[DEF:Project_Map]`
-*   **Apache Superset OpenAPI:** `.ai/openapi.json` -> `[DEF:Doc:Superset_OpenAPI]`
+*   **High-level Module Map:** `.ai/structure/MODULE_MAP.md` -> `[DEF:Module_Map]`
+*   **Low-level Project Map:** `.ai/structure/PROJECT_MAP.md` -> `[DEF:Project_Map]`
+*   **Apache Superset OpenAPI:** `.ai/openapi/superset_openapi.json` -> `[DEF:Doc:Superset_OpenAPI]`
 *   **Backend Core:** `backend/src/core` -> `[DEF:Module:Backend_Core]`
 *   **Backend API:** `backend/src/api` -> `[DEF:Module:Backend_API]`
 *   **Frontend Lib:** `frontend/src/lib` -> `[DEF:Module:Frontend_Lib]`
--- a/.ai/openapi/superset_openapi.json
+++ b/.ai/openapi/superset_openapi.json
--- a/.ai/structure/MODULE_MAP.md
+++ b/.ai/structure/MODULE_MAP.md
--- a/.ai/structure/PROJECT_MAP.md
+++ b/.ai/structure/PROJECT_MAP.md
--- a/backend/src/api/routes/tests/test_dashboards.py
+++ b/backend/src/api/routes/tests/test_dashboards.py
@@ -6,6 +6,7 @@

 import pytest
 from unittest.mock import MagicMock, patch, AsyncMock
+from datetime import datetime, timezone
 from fastapi.testclient import TestClient
 from src.app import app
 from src.api.routes.dashboards import DashboardsResponse
@@ -354,4 +355,84 @@ def test_get_database_mappings_success():
 # [/DEF:test_get_database_mappings_success:Function]


+# [DEF:test_get_dashboard_tasks_history_filters_success:Function]
+# @TEST: GET /api/dashboards/{id}/tasks returns backup and llm tasks for dashboard
+def test_get_dashboard_tasks_history_filters_success():
+    with patch("src.api.routes.dashboards.get_task_manager") as mock_task_mgr, \
+         patch("src.api.routes.dashboards.has_permission") as mock_perm:
+        now = datetime.now(timezone.utc)
+
+        llm_task = MagicMock()
+        llm_task.id = "task-llm-1"
+        llm_task.plugin_id = "llm_dashboard_validation"
+        llm_task.status = "SUCCESS"
+        llm_task.started_at = now
+        llm_task.finished_at = now
+        llm_task.params = {"dashboard_id": "42", "environment_id": "prod"}
+        llm_task.result = {"summary": "LLM validation complete"}
+
+        backup_task = MagicMock()
+        backup_task.id = "task-backup-1"
+        backup_task.plugin_id = "superset-backup"
+        backup_task.status = "RUNNING"
+        backup_task.started_at = now
+        backup_task.finished_at = None
+        backup_task.params = {"env": "prod", "dashboards": [42]}
+        backup_task.result = {}
+
+        other_task = MagicMock()
+        other_task.id = "task-other"
+        other_task.plugin_id = "superset-backup"
+        other_task.status = "SUCCESS"
+        other_task.started_at = now
+        other_task.finished_at = now
+        other_task.params = {"env": "prod", "dashboards": [777]}
+        other_task.result = {}
+
+        mock_task_mgr.return_value.get_all_tasks.return_value = [other_task, llm_task, backup_task]
+        mock_perm.return_value = lambda: True
+
+        response = client.get("/api/dashboards/42/tasks?env_id=prod&limit=10")
+
+        assert response.status_code == 200
+        data = response.json()
+        assert data["dashboard_id"] == 42
+        assert len(data["items"]) == 2
+        assert {item["plugin_id"] for item in data["items"]} == {"llm_dashboard_validation", "superset-backup"}
+# [/DEF:test_get_dashboard_tasks_history_filters_success:Function]
+
+
+# [DEF:test_get_dashboard_thumbnail_success:Function]
+# @TEST: GET /api/dashboards/{id}/thumbnail proxies image bytes from Superset
+def test_get_dashboard_thumbnail_success():
+    with patch("src.api.routes.dashboards.get_config_manager") as mock_config, \
+         patch("src.api.routes.dashboards.has_permission") as mock_perm, \
+         patch("src.api.routes.dashboards.SupersetClient") as mock_client_cls:
+        mock_env = MagicMock()
+        mock_env.id = "prod"
+        mock_config.return_value.get_environments.return_value = [mock_env]
+        mock_perm.return_value = lambda: True
+
+        mock_client = MagicMock()
+        mock_response = MagicMock()
+        mock_response.status_code = 200
+        mock_response.content = b"fake-image-bytes"
+        mock_response.headers = {"Content-Type": "image/png"}
+
+        def _network_request(method, endpoint, **kwargs):
+            if method == "POST":
+                return {"image_url": "/api/v1/dashboard/42/screenshot/abc123/"}
+            return mock_response
+
+        mock_client.network.request.side_effect = _network_request
+        mock_client_cls.return_value = mock_client
+
+        response = client.get("/api/dashboards/42/thumbnail?env_id=prod")
+
+        assert response.status_code == 200
+        assert response.content == b"fake-image-bytes"
+        assert response.headers["content-type"].startswith("image/png")
+# [/DEF:test_get_dashboard_thumbnail_success:Function]
+
+
 # [/DEF:backend.src.api.routes.__tests__.test_dashboards:Module]
--- a/backend/src/api/routes/dashboards.py
+++ b/backend/src/api/routes/dashboards.py
@@ -11,12 +11,16 @@
 # @INVARIANT: All dashboard responses include git_status and last_task metadata

 # [SECTION: IMPORTS]
-from fastapi import APIRouter, Depends, HTTPException
-from typing import List, Optional, Dict
+from fastapi import APIRouter, Depends, HTTPException, Query, Response
+from fastapi.responses import JSONResponse
+from typing import List, Optional, Dict, Any
+import re
+from urllib.parse import urlparse
 from pydantic import BaseModel, Field
 from ...dependencies import get_config_manager, get_task_manager, get_resource_service, get_mapping_service, has_permission
 from ...core.logger import logger, belief_scope
 from ...core.superset_client import SupersetClient
+from ...core.utils.network import DashboardNotFoundError
 # [/SECTION]

 router = APIRouter(prefix="/api/dashboards", tags=["Dashboards"])
@@ -90,6 +94,24 @@ class DashboardDetailResponse(BaseModel):
    dataset_count: int
 # [/DEF:DashboardDetailResponse:DataClass]

+# [DEF:DashboardTaskHistoryItem:DataClass]
+class DashboardTaskHistoryItem(BaseModel):
+    id: str
+    plugin_id: str
+    status: str
+    validation_status: Optional[str] = None
+    started_at: Optional[str] = None
+    finished_at: Optional[str] = None
+    env_id: Optional[str] = None
+    summary: Optional[str] = None
+# [/DEF:DashboardTaskHistoryItem:DataClass]
+
+# [DEF:DashboardTaskHistoryResponse:DataClass]
+class DashboardTaskHistoryResponse(BaseModel):
+    dashboard_id: int
+    items: List[DashboardTaskHistoryItem]
+# [/DEF:DashboardTaskHistoryResponse:DataClass]
+
 # [DEF:DatabaseMapping:DataClass]
 class DatabaseMapping(BaseModel):
    source_db: str
@@ -259,6 +281,190 @@ async def get_dashboard_detail(
            raise HTTPException(status_code=503, detail=f"Failed to fetch dashboard detail: {str(e)}")
 # [/DEF:get_dashboard_detail:Function]

+
+# [DEF:_task_matches_dashboard:Function]
+# @PURPOSE: Checks whether task params are tied to a specific dashboard and environment.
+# @PRE: task-like object exposes plugin_id and params fields.
+# @POST: Returns True only for supported task plugins tied to dashboard_id (+optional env_id).
+def _task_matches_dashboard(task: Any, dashboard_id: int, env_id: Optional[str]) -> bool:
+    plugin_id = getattr(task, "plugin_id", None)
+    if plugin_id not in {"superset-backup", "llm_dashboard_validation"}:
+        return False
+
+    params = getattr(task, "params", {}) or {}
+    dashboard_id_str = str(dashboard_id)
+
+    if plugin_id == "llm_dashboard_validation":
+        task_dashboard_id = params.get("dashboard_id")
+        if str(task_dashboard_id) != dashboard_id_str:
+            return False
+        if env_id:
+            task_env = params.get("environment_id")
+            return str(task_env) == str(env_id)
+        return True
+
+    # superset-backup can pass dashboards as "dashboard_ids" or "dashboards"
+    dashboard_ids = params.get("dashboard_ids") or params.get("dashboards") or []
+    normalized_ids = {str(item) for item in dashboard_ids}
+    if dashboard_id_str not in normalized_ids:
+        return False
+    if env_id:
+        task_env = params.get("environment_id") or params.get("env")
+        return str(task_env) == str(env_id)
+    return True
+# [/DEF:_task_matches_dashboard:Function]
+
+
+# [DEF:get_dashboard_tasks_history:Function]
+# @PURPOSE: Returns history of backup and LLM validation tasks for a dashboard.
+# @PRE: dashboard_id is valid integer.
+# @POST: Response contains sorted task history (newest first).
+@router.get("/{dashboard_id:int}/tasks", response_model=DashboardTaskHistoryResponse)
+async def get_dashboard_tasks_history(
+    dashboard_id: int,
+    env_id: Optional[str] = None,
+    limit: int = Query(20, ge=1, le=100),
+    task_manager=Depends(get_task_manager),
+    _ = Depends(has_permission("tasks", "READ"))
+):
+    with belief_scope("get_dashboard_tasks_history", f"dashboard_id={dashboard_id}, env_id={env_id}, limit={limit}"):
+        matching_tasks = []
+        for task in task_manager.get_all_tasks():
+            if _task_matches_dashboard(task, dashboard_id, env_id):
+                matching_tasks.append(task)
+
+        def _sort_key(task_obj: Any) -> str:
+            return (
+                str(getattr(task_obj, "started_at", "") or "")
+                or str(getattr(task_obj, "finished_at", "") or "")
+            )
+
+        matching_tasks.sort(key=_sort_key, reverse=True)
+        selected = matching_tasks[:limit]
+
+        items = []
+        for task in selected:
+            result = getattr(task, "result", None)
+            summary = None
+            validation_status = None
+            if isinstance(result, dict):
+                raw_validation_status = result.get("status")
+                if raw_validation_status is not None:
+                    validation_status = str(raw_validation_status)
+                summary = (
+                    result.get("summary")
+                    or result.get("status")
+                    or result.get("message")
+                )
+            params = getattr(task, "params", {}) or {}
+            items.append(
+                DashboardTaskHistoryItem(
+                    id=str(getattr(task, "id", "")),
+                    plugin_id=str(getattr(task, "plugin_id", "")),
+                    status=str(getattr(task, "status", "")),
+                    validation_status=validation_status,
+                    started_at=getattr(task, "started_at", None).isoformat() if getattr(task, "started_at", None) else None,
+                    finished_at=getattr(task, "finished_at", None).isoformat() if getattr(task, "finished_at", None) else None,
+                    env_id=str(params.get("environment_id") or params.get("env")) if (params.get("environment_id") or params.get("env")) else None,
+                    summary=summary,
+                )
+            )
+
+        logger.info(f"[get_dashboard_tasks_history][Coherence:OK] Found {len(items)} tasks for dashboard {dashboard_id}")
+        return DashboardTaskHistoryResponse(dashboard_id=dashboard_id, items=items)
+# [/DEF:get_dashboard_tasks_history:Function]
+
+
+# [DEF:get_dashboard_thumbnail:Function]
+# @PURPOSE: Proxies Superset dashboard thumbnail with cache support.
+# @PRE: env_id must exist.
+# @POST: Returns image bytes or 202 when thumbnail is being prepared by Superset.
+@router.get("/{dashboard_id:int}/thumbnail")
+async def get_dashboard_thumbnail(
+    dashboard_id: int,
+    env_id: str,
+    force: bool = Query(False),
+    config_manager=Depends(get_config_manager),
+    _ = Depends(has_permission("plugin:migration", "READ"))
+):
+    with belief_scope("get_dashboard_thumbnail", f"dashboard_id={dashboard_id}, env_id={env_id}, force={force}"):
+        environments = config_manager.get_environments()
+        env = next((e for e in environments if e.id == env_id), None)
+        if not env:
+            logger.error(f"[get_dashboard_thumbnail][Coherence:Failed] Environment not found: {env_id}")
+            raise HTTPException(status_code=404, detail="Environment not found")
+
+        try:
+            client = SupersetClient(env)
+            digest = None
+            thumb_endpoint = None
+
+            # Preferred flow (newer Superset): ask server to cache screenshot and return digest/image_url.
+            try:
+                screenshot_payload = client.network.request(
+                    method="POST",
+                    endpoint=f"/dashboard/{dashboard_id}/cache_dashboard_screenshot/",
+                    json={"force": force},
+                )
+                payload = screenshot_payload.get("result", screenshot_payload) if isinstance(screenshot_payload, dict) else {}
+                image_url = payload.get("image_url", "") if isinstance(payload, dict) else ""
+                if isinstance(image_url, str) and image_url:
+                    matched = re.search(r"/dashboard/\d+/(?:thumbnail|screenshot)/([^/]+)/?$", image_url)
+                    if matched:
+                        digest = matched.group(1)
+            except DashboardNotFoundError:
+                logger.warning(
+                    "[get_dashboard_thumbnail][Fallback] cache_dashboard_screenshot endpoint unavailable, fallback to dashboard.thumbnail_url"
+                )
+
+            # Fallback flow (older Superset): read thumbnail_url from dashboard payload.
+            if not digest:
+                dashboard_payload = client.network.request(
+                    method="GET",
+                    endpoint=f"/dashboard/{dashboard_id}",
+                )
+                dashboard_data = dashboard_payload.get("result", dashboard_payload) if isinstance(dashboard_payload, dict) else {}
+                thumbnail_url = dashboard_data.get("thumbnail_url", "") if isinstance(dashboard_data, dict) else ""
+                if isinstance(thumbnail_url, str) and thumbnail_url:
+                    parsed = urlparse(thumbnail_url)
+                    parsed_path = parsed.path or thumbnail_url
+                    if parsed_path.startswith("/api/v1/"):
+                        parsed_path = parsed_path[len("/api/v1"):]
+                    thumb_endpoint = parsed_path
+                    matched = re.search(r"/dashboard/\d+/(?:thumbnail|screenshot)/([^/]+)/?$", parsed_path)
+                    if matched:
+                        digest = matched.group(1)
+
+            if not thumb_endpoint:
+                thumb_endpoint = f"/dashboard/{dashboard_id}/thumbnail/{digest or 'latest'}/"
+
+            thumb_response = client.network.request(
+                method="GET",
+                endpoint=thumb_endpoint,
+                raw_response=True,
+                allow_redirects=True,
+            )
+
+            if thumb_response.status_code == 202:
+                payload_202: Dict[str, Any] = {}
+                try:
+                    payload_202 = thumb_response.json()
+                except Exception:
+                    payload_202 = {"message": "Thumbnail is being generated"}
+                return JSONResponse(status_code=202, content=payload_202)
+
+            content_type = thumb_response.headers.get("Content-Type", "image/png")
+            return Response(content=thumb_response.content, media_type=content_type)
+        except DashboardNotFoundError as e:
+            logger.error(f"[get_dashboard_thumbnail][Coherence:Failed] Dashboard not found for thumbnail: {e}")
+            raise HTTPException(status_code=404, detail="Dashboard thumbnail not found")
+        except HTTPException:
+            raise
+        except Exception as e:
+            logger.error(f"[get_dashboard_thumbnail][Coherence:Failed] Failed to fetch dashboard thumbnail: {e}")
+            raise HTTPException(status_code=503, detail=f"Failed to fetch dashboard thumbnail: {str(e)}")
+# [/DEF:get_dashboard_thumbnail:Function]
+
 # [DEF:MigrateRequest:DataClass]
 class MigrateRequest(BaseModel):
    source_env_id: str = Field(..., description="Source environment ID")
--- a/backend/src/api/routes/llm.py
+++ b/backend/src/api/routes/llm.py
@@ -5,7 +5,7 @@
 # @LAYER: UI (API)

 from fastapi import APIRouter, Depends, HTTPException, status
-from typing import List
+from typing import List, Optional
 from ...core.logger import logger
 from ...schemas.auth import User
 from ...dependencies import get_current_user as get_current_active_user
@@ -19,6 +19,20 @@ from sqlalchemy.orm import Session
 router = APIRouter(tags=["LLM"])
 # [/DEF:router:Global]

+
+# [DEF:_is_valid_runtime_api_key:Function]
+# @PURPOSE: Validate decrypted runtime API key presence/shape.
+# @PRE: value can be None.
+# @POST: Returns True only for non-placeholder key.
+def _is_valid_runtime_api_key(value: Optional[str]) -> bool:
+    key = (value or "").strip()
+    if not key:
+        return False
+    if key in {"********", "EMPTY_OR_NONE"}:
+        return False
+    return len(key) >= 16
+# [/DEF:_is_valid_runtime_api_key:Function]
+
 # [DEF:get_providers:Function]
 # @PURPOSE: Retrieve all LLM provider configurations.
 # @PRE: User is authenticated.
@@ -47,6 +61,37 @@ async def get_providers(
    ]
 # [/DEF:get_providers:Function]

+
+# [DEF:get_llm_status:Function]
+# @PURPOSE: Returns whether LLM runtime is configured for dashboard validation.
+# @PRE: User is authenticated.
+# @POST: configured=true only when an active provider with valid decrypted key exists.
+@router.get("/status")
+async def get_llm_status(
+    current_user: User = Depends(get_current_active_user),
+    db: Session = Depends(get_db)
+):
+    service = LLMProviderService(db)
+    providers = service.get_all_providers()
+    active_provider = next((p for p in providers if p.is_active), None)
+
+    if not active_provider:
+        return {"configured": False, "reason": "no_active_provider"}
+
+    api_key = service.get_decrypted_api_key(active_provider.id)
+    if not _is_valid_runtime_api_key(api_key):
+        return {"configured": False, "reason": "invalid_api_key"}
+
+    return {
+        "configured": True,
+        "reason": "ok",
+        "provider_id": active_provider.id,
+        "provider_name": active_provider.name,
+        "provider_type": active_provider.provider_type,
+        "default_model": active_provider.default_model,
+    }
+# [/DEF:get_llm_status:Function]
+
 # [DEF:create_provider:Function]
 # @PURPOSE: Create a new LLM provider configuration.
 # @PRE: User is authenticated and has admin permissions.
--- a/backend/src/api/routes/storage.py
+++ b/backend/src/api/routes/storage.py
@@ -144,4 +144,46 @@ async def download_file(
            raise HTTPException(status_code=400, detail=str(e))
 # [/DEF:download_file:Function]

+# [DEF:get_file_by_path:Function]
+# @PURPOSE: Retrieve a file by validated absolute/relative path under storage root.
+#
+# @PRE: path must resolve under configured storage root.
+# @POST: Returns a FileResponse for existing files.
+#
+# @PARAM: path (str) - Absolute or storage-root-relative file path.
+# @RETURN: FileResponse - The file content.
+#
+# @RELATION:    CALLS -> StoragePlugin.get_storage_root
+# @RELATION:    CALLS -> StoragePlugin.validate_path
+@router.get("/file")
+async def get_file_by_path(
+    path: str,
+    plugin_loader=Depends(get_plugin_loader),
+    _ = Depends(has_permission("plugin:storage", "READ"))
+):
+    with belief_scope("get_file_by_path"):
+        storage_plugin: StoragePlugin = plugin_loader.get_plugin("storage-manager")
+        if not storage_plugin:
+            raise HTTPException(status_code=500, detail="Storage plugin not loaded")
+
+        requested_path = (path or "").strip()
+        if not requested_path:
+            raise HTTPException(status_code=400, detail="Path is required")
+
+        try:
+            candidate = Path(requested_path)
+            if candidate.is_absolute():
+                abs_path = storage_plugin.validate_path(candidate)
+            else:
+                storage_root = storage_plugin.get_storage_root()
+                abs_path = storage_plugin.validate_path(storage_root / candidate)
+        except ValueError as e:
+            raise HTTPException(status_code=400, detail=str(e))
+
+        if not abs_path.exists() or not abs_path.is_file():
+            raise HTTPException(status_code=404, detail="File not found")
+
+        return FileResponse(path=str(abs_path), filename=abs_path.name)
+# [/DEF:get_file_by_path:Function]
+
 # [/DEF:storage_routes:Module]
--- a/backend/src/plugins/llm_analysis/plugin.py
+++ b/backend/src/plugins/llm_analysis/plugin.py
@@ -30,6 +30,34 @@ from ...services.llm_prompt_templates import (
    render_prompt,
 )

+# [DEF:_is_masked_or_invalid_api_key:Function]
+# @PURPOSE: Guards against placeholder or malformed API keys in runtime.
+# @PRE: value may be None.
+# @POST: Returns True when value cannot be used for authenticated provider calls.
+def _is_masked_or_invalid_api_key(value: Optional[str]) -> bool:
+    key = (value or "").strip()
+    if not key:
+        return True
+    if key in {"********", "EMPTY_OR_NONE"}:
+        return True
+    # Most provider tokens are significantly longer; short values are almost always placeholders.
+    return len(key) < 16
+# [/DEF:_is_masked_or_invalid_api_key:Function]
+
+# [DEF:_json_safe_value:Function]
+# @PURPOSE: Recursively normalize payload values for JSON serialization.
+# @PRE: value may be nested dict/list with datetime values.
+# @POST: datetime values are converted to ISO strings.
+def _json_safe_value(value: Any):
+    if isinstance(value, datetime):
+        return value.isoformat()
+    if isinstance(value, dict):
+        return {k: _json_safe_value(v) for k, v in value.items()}
+    if isinstance(value, list):
+        return [_json_safe_value(v) for v in value]
+    return value
+# [/DEF:_json_safe_value:Function]
+
 # [DEF:DashboardValidationPlugin:Class]
 # @PURPOSE: Plugin for automated dashboard health analysis using LLMs.
 # @RELATION: IMPLEMENTS -> backend.src.core.plugin_base.PluginBase
@@ -70,6 +98,7 @@ class DashboardValidationPlugin(PluginBase):
    # @SIDE_EFFECT: Captures a screenshot, calls LLM API, and writes to the database.
    async def execute(self, params: Dict[str, Any], context: Optional[TaskContext] = None):
        with belief_scope("execute", f"plugin_id={self.id}"):
+            validation_started_at = datetime.utcnow()
            # Use TaskContext logger if available, otherwise fall back to app logger
            log = context.logger if context else logger
            
@@ -118,11 +147,10 @@ class DashboardValidationPlugin(PluginBase):
                llm_log.debug(f"API Key decrypted (first 8 chars): {api_key[:8] if api_key and len(api_key) > 8 else 'EMPTY_OR_NONE'}...")
                
                # Check if API key was successfully decrypted
-                if not api_key:
+                if _is_masked_or_invalid_api_key(api_key):
                    raise ValueError(
-                        f"Failed to decrypt API key for provider {provider_id}. "
-                        f"The provider may have been encrypted with a different encryption key. "
-                        f"Please update the provider with a new API key through the UI."
+                        f"Invalid API key for provider {provider_id}. "
+                        "Please open LLM provider settings and save a real API key (not masked placeholder)."
                    )

                # 3. Capture Screenshot
@@ -135,12 +163,15 @@ class DashboardValidationPlugin(PluginBase):
                filename = f"{dashboard_id}_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png"
                screenshot_path = os.path.join(screenshots_dir, filename)
                
+                screenshot_started_at = datetime.utcnow()
                screenshot_log.info(f"Capturing screenshot for dashboard {dashboard_id}")
                await screenshot_service.capture_dashboard(dashboard_id, screenshot_path)
                screenshot_log.debug(f"Screenshot saved to: {screenshot_path}")
+                screenshot_finished_at = datetime.utcnow()

                # 4. Fetch Logs (from Environment /api/v1/log/)
                logs = []
+                logs_fetch_started_at = datetime.utcnow()
                try:
                    client = SupersetClient(env)
                    
@@ -181,6 +212,7 @@ class DashboardValidationPlugin(PluginBase):
                except Exception as e:
                    superset_log.warning(f"Failed to fetch logs from environment: {e}")
                    logs = [f"Error fetching remote logs: {str(e)}"]
+                logs_fetch_finished_at = datetime.utcnow()

                # 5. Analyze with LLM
                llm_client = LLMClient(
@@ -196,11 +228,13 @@ class DashboardValidationPlugin(PluginBase):
                    "dashboard_validation_prompt",
                    DEFAULT_LLM_PROMPTS["dashboard_validation_prompt"],
                )
+                llm_call_started_at = datetime.utcnow()
                analysis = await llm_client.analyze_dashboard(
                    screenshot_path,
                    logs,
                    prompt_template=dashboard_prompt,
                )
+                llm_call_finished_at = datetime.utcnow()
                
                # Log analysis summary to task logs for better visibility
                llm_log.info(f"[ANALYSIS_SUMMARY] Status: {analysis['status']}")
@@ -218,6 +252,35 @@ class DashboardValidationPlugin(PluginBase):
                    screenshot_path=screenshot_path,
                    raw_response=str(analysis)
                )
+                validation_finished_at = datetime.utcnow()
+
+                result_payload = _json_safe_value(validation_result.dict())
+                result_payload["screenshot_paths"] = [screenshot_path]
+                result_payload["logs_sent_to_llm"] = logs
+                result_payload["logs_sent_count"] = len(logs)
+                result_payload["prompt_template"] = dashboard_prompt
+                result_payload["provider"] = {
+                    "id": db_provider.id,
+                    "name": db_provider.name,
+                    "type": db_provider.provider_type,
+                    "base_url": db_provider.base_url,
+                    "model": db_provider.default_model,
+                }
+                result_payload["environment_id"] = env_id
+                result_payload["timings"] = {
+                    "validation_started_at": validation_started_at.isoformat(),
+                    "validation_finished_at": validation_finished_at.isoformat(),
+                    "validation_duration_ms": int((validation_finished_at - validation_started_at).total_seconds() * 1000),
+                    "screenshot_started_at": screenshot_started_at.isoformat(),
+                    "screenshot_finished_at": screenshot_finished_at.isoformat(),
+                    "screenshot_duration_ms": int((screenshot_finished_at - screenshot_started_at).total_seconds() * 1000),
+                    "logs_fetch_started_at": logs_fetch_started_at.isoformat(),
+                    "logs_fetch_finished_at": logs_fetch_finished_at.isoformat(),
+                    "logs_fetch_duration_ms": int((logs_fetch_finished_at - logs_fetch_started_at).total_seconds() * 1000),
+                    "llm_call_started_at": llm_call_started_at.isoformat(),
+                    "llm_call_finished_at": llm_call_finished_at.isoformat(),
+                    "llm_call_duration_ms": int((llm_call_finished_at - llm_call_started_at).total_seconds() * 1000),
+                }

                db_record = ValidationRecord(
                    dashboard_id=validation_result.dashboard_id,
@@ -225,7 +288,7 @@ class DashboardValidationPlugin(PluginBase):
                    summary=validation_result.summary,
                    issues=[issue.dict() for issue in validation_result.issues],
                    screenshot_path=validation_result.screenshot_path,
-                    raw_response=validation_result.raw_response
+                    raw_response=json.dumps(result_payload, ensure_ascii=False)
                )
                db.add(db_record)
                db.commit()
@@ -240,7 +303,7 @@ class DashboardValidationPlugin(PluginBase):
                # Final log to ensure all analysis is visible in task logs
                log.info(f"Validation completed for dashboard {dashboard_id}. Status: {validation_result.status.value}")
                
-                return validation_result.dict()
+                return result_payload

            finally:
                db.close()
@@ -328,11 +391,10 @@ class DocumentationPlugin(PluginBase):
                llm_log.debug(f"API Key decrypted (first 8 chars): {api_key[:8] if api_key and len(api_key) > 8 else 'EMPTY_OR_NONE'}...")
                
                # Check if API key was successfully decrypted
-                if not api_key:
+                if _is_masked_or_invalid_api_key(api_key):
                    raise ValueError(
-                        f"Failed to decrypt API key for provider {provider_id}. "
-                        f"The provider may have been encrypted with a different encryption key. "
-                        f"Please update the provider with a new API key through the UI."
+                        f"Invalid API key for provider {provider_id}. "
+                        "Please open LLM provider settings and save a real API key (not masked placeholder)."
                    )

                # 3. Fetch Metadata (US2 / T024)
--- a/backend/src/plugins/llm_analysis/service.py
+++ b/backend/src/plugins/llm_analysis/service.py
@@ -13,6 +13,7 @@ import base64
 import json
 import io
 from typing import List, Dict, Any
+import httpx
 from PIL import Image
 from playwright.async_api import async_playwright
 from openai import AsyncOpenAI, RateLimitError, AuthenticationError as OpenAIAuthenticationError
@@ -422,7 +423,10 @@ class LLMClient:
    # @PRE: api_key, base_url, and default_model are non-empty strings.
    def __init__(self, provider_type: LLMProviderType, api_key: str, base_url: str, default_model: str):
        self.provider_type = provider_type
-        self.api_key = api_key
+        normalized_key = (api_key or "").strip()
+        if normalized_key.lower().startswith("bearer "):
+            normalized_key = normalized_key[7:].strip()
+        self.api_key = normalized_key
        self.base_url = base_url
        self.default_model = default_model
        
@@ -431,10 +435,22 @@ class LLMClient:
        logger.info(f"[LLMClient.__init__]   Provider Type: {provider_type}")
        logger.info(f"[LLMClient.__init__]   Base URL: {base_url}")
        logger.info(f"[LLMClient.__init__]   Default Model: {default_model}")
-        logger.info(f"[LLMClient.__init__]   API Key (first 8 chars): {api_key[:8] if api_key and len(api_key) > 8 else 'EMPTY_OR_NONE'}...")
-        logger.info(f"[LLMClient.__init__]   API Key Length: {len(api_key) if api_key else 0}")
+        logger.info(f"[LLMClient.__init__]   API Key (first 8 chars): {self.api_key[:8] if self.api_key and len(self.api_key) > 8 else 'EMPTY_OR_NONE'}...")
+        logger.info(f"[LLMClient.__init__]   API Key Length: {len(self.api_key) if self.api_key else 0}")

-        self.client = AsyncOpenAI(api_key=api_key, base_url=base_url)
+        # Some OpenAI-compatible gateways are strict about auth header naming.
+        default_headers = {"Authorization": f"Bearer {self.api_key}"}
+        if self.provider_type == LLMProviderType.KILO:
+            default_headers["Authentication"] = f"Bearer {self.api_key}"
+            default_headers["X-API-Key"] = self.api_key
+
+        http_client = httpx.AsyncClient(headers=default_headers, timeout=120.0)
+        self.client = AsyncOpenAI(
+            api_key=self.api_key,
+            base_url=base_url,
+            default_headers=default_headers,
+            http_client=http_client,
+        )
    # [/DEF:LLMClient.__init__:Function]

    # [DEF:LLMClient._supports_json_response_format:Function]
--- a/check_test_data.py
+++ b/check_test_data.py
@@ -0,0 +1,27 @@
+import os
+
+def check_file(filepath):
+    try:
+        with open(filepath, 'r', encoding='utf-8') as f:
+            content = f.read()
+            if '@TIER: CRITICAL' in content:
+                if '@TEST_DATA' not in content:
+                    return filepath
+    except Exception as e:
+        print(f"Error reading {filepath}: {e}")
+    return None
+
+missing_files = []
+for root_dir in ['backend/src', 'frontend/src']:
+    for dirpath, _, filenames in os.walk(root_dir):
+        for name in filenames:
+            ext = os.path.splitext(name)[1]
+            if ext in ['.py', '.js', '.ts', '.svelte']:
+                full_path = os.path.join(dirpath, name)
+                res = check_file(full_path)
+                if res:
+                    missing_files.append(res)
+
+print("Files missing @TEST_DATA:")
+for f in missing_files:
+    print(f)
--- a/dashboards_example/dashboard_export_20260114T112108/charts/FI-0083_4019.yaml
+++ b/dashboards_example/dashboard_export_20260114T112108/charts/FI-0083_4019.yaml
@@ -1,55 +0,0 @@
-slice_name: "FI-0083 \u0421\u0442\u0430\u0442\u0438\u0441\u0442\u0438\u043A\u0430\
-  \ \u043F\u043E \u0414\u0417/\u041F\u0414\u0417"
-description: null
-certified_by: null
-certification_details: null
-viz_type: pivot_table_v2
-params:
-  datasource: 859__table
-  viz_type: pivot_table_v2
-  slice_id: 4019
-  groupbyColumns:
-  - dt
-  groupbyRows:
-  - counterparty_search_name
-  - attribute
-  time_grain_sqla: P1M
-  temporal_columns_lookup:
-    dt: true
-  metrics:
-  - m_debt_amount
-  - m_overdue_amount
-  metricsLayout: COLUMNS
-  adhoc_filters:
-  - clause: WHERE
-    comparator: No filter
-    expressionType: SIMPLE
-    operator: TEMPORAL_RANGE
-    subject: dt
-  row_limit: '90000'
-  order_desc: false
-  aggregateFunction: Sum
-  combineMetric: true
-  valueFormat: SMART_NUMBER
-  date_format: smart_date
-  rowOrder: key_a_to_z
-  colOrder: key_a_to_z
-  value_font_size: 12
-  header_font_size: 12
-  label_align: left
-  column_config:
-    m_debt_amount:
-      d3NumberFormat: ',d'
-    m_overdue_amount:
-      d3NumberFormat: ',d'
-  conditional_formatting: []
-  extra_form_data: {}
-  dashboards:
-  - 184
-query_context: '{"datasource":{"id":859,"type":"table"},"force":false,"queries":[{"filters":[{"col":"dt","op":"TEMPORAL_RANGE","val":"No
-  filter"}],"extras":{"having":"","where":""},"applied_time_extras":{},"columns":[{"timeGrain":"P1M","columnType":"BASE_AXIS","sqlExpression":"dt","label":"dt","expressionType":"SQL"},"counterparty_search_name","attribute"],"metrics":["m_debt_amount","m_overdue_amount"],"orderby":[["m_debt_amount",true]],"annotation_layers":[],"row_limit":90000,"series_limit":0,"order_desc":false,"url_params":{},"custom_params":{},"custom_form_data":{}}],"form_data":{"datasource":"859__table","viz_type":"pivot_table_v2","slice_id":4019,"groupbyColumns":["dt"],"groupbyRows":["counterparty_search_name","attribute"],"time_grain_sqla":"P1M","temporal_columns_lookup":{"dt":true},"metrics":["m_debt_amount","m_overdue_amount"],"metricsLayout":"COLUMNS","adhoc_filters":[{"clause":"WHERE","comparator":"No
-  filter","expressionType":"SIMPLE","operator":"TEMPORAL_RANGE","subject":"dt"}],"row_limit":"90000","order_desc":false,"aggregateFunction":"Sum","combineMetric":true,"valueFormat":"SMART_NUMBER","date_format":"smart_date","rowOrder":"key_a_to_z","colOrder":"key_a_to_z","value_font_size":12,"header_font_size":12,"label_align":"left","column_config":{"m_debt_amount":{"d3NumberFormat":",d"},"m_overdue_amount":{"d3NumberFormat":",d"}},"conditional_formatting":[],"extra_form_data":{},"dashboards":[184],"force":false,"result_format":"json","result_type":"full"},"result_format":"json","result_type":"full"}'
-cache_timeout: null
-uuid: 9c293065-73e2-4d9b-a175-d188ff8ef575
-version: 1.0.0
-dataset_uuid: 9e645dc0-da25-4f61-9465-6e649b0bc4b1
--- a/dashboards_example/dashboard_export_20260114T112108/databases/Prod_Clickhouse_10.yaml
+++ b/dashboards_example/dashboard_export_20260114T112108/databases/Prod_Clickhouse_10.yaml
@@ -1,13 +0,0 @@
-database_name: Prod Clickhouse
-sqlalchemy_uri: clickhousedb+connect://viz_superset_click_prod:XXXXXXXXXX@rgm-s-khclk.hq.root.ad:443/dm
-cache_timeout: null
-expose_in_sqllab: true
-allow_run_async: false
-allow_ctas: false
-allow_cvas: false
-allow_dml: true
-allow_file_upload: false
-extra:
-  allows_virtual_table_explore: true
-uuid: 97aced68-326a-4094-b381-27980560efa9
-version: 1.0.0
--- a/dashboards_example/dashboard_export_20260114T112108/datasets/Prod_Clickhouse_10/FI-0080-06_875.yaml
+++ b/dashboards_example/dashboard_export_20260114T112108/datasets/Prod_Clickhouse_10/FI-0080-06_875.yaml
@@ -1,119 +0,0 @@
-table_name: "FI-0080-06 \u041A\u0430\u043B\u0435\u043D\u0434\u0430\u0440\u044C (\u041E\
-  \u0431\u0449\u0438\u0439 \u0441\u043F\u0440\u0430\u0432\u043E\u0447\u043D\u0438\u043A\
-  )"
-main_dttm_col: null
-description: null
-default_endpoint: null
-offset: 0
-cache_timeout: null
-schema: dm_view
-sql: "-- [HEADER]\r\n-- [\u041D\u0410\u0417\u041D\u0410\u0427\u0415\u041D\u0418\u0415\
-  ]: \u041F\u043E\u043B\u0443\u0447\u0435\u043D\u0438\u0435 \u0434\u0438\u0430\u043F\
-  \u0430\u0437\u043E\u043D\u0430 \u0434\u0430\u0442 \u0434\u043B\u044F \u043E\u0442\
-  \u0447\u0435\u0442\u0430 \u043E \u0437\u0430\u0434\u043E\u043B\u0436\u0435\u043D\
-  \u043D\u043E\u0441\u0442\u044F\u0445 \u043F\u043E \u043E\u0431\u043E\u0440\u043E\
-  \u0442\u043D\u044B\u043C \u0441\u0440\u0435\u0434\u0441\u0442\u0432\u0430\u043C\r\
-  \n-- [\u041A\u041B\u042E\u0427\u0415\u0412\u042B\u0415 \u041A\u041E\u041B\u041E\u041D\
-  \u041A\u0418]:\r\n--   - from_dt_txt: \u041D\u0430\u0447\u0430\u043B\u044C\u043D\
-  \u0430\u044F \u0434\u0430\u0442\u0430 \u0432 \u0444\u043E\u0440\u043C\u0430\u0442\
-  \u0435 DD.MM.YYYY\r\n--   - to_dt_txt: \u041A\u043E\u043D\u0435\u0447\u043D\u0430\
-  \u044F \u0434\u0430\u0442\u0430 \u0432 \u0444\u043E\u0440\u043C\u0430\u0442\u0435\
-  \ DD.MM.YYYY\r\n-- [JINJA \u041F\u0410\u0420\u0410\u041C\u0415\u0422\u0420\u042B\
-  ]:\r\n--   - {{ filter_values(\"yes_no_check\") }}: \u0424\u0438\u043B\u044C\u0442\
-  \u0440 \"\u0414\u0430/\u041D\u0435\u0442\" \u0434\u043B\u044F \u043E\u0433\u0440\
-  \u0430\u043D\u0438\u0447\u0435\u043D\u0438\u044F \u0432\u044B\u0431\u043E\u0440\u043A\
-  \u0438 \u043F\u043E \u0434\u0430\u0442\u0435\r\n-- [\u041B\u041E\u0413\u0418\u041A\
-  \u0410]: \u041E\u043F\u0440\u0435\u0434\u0435\u043B\u044F\u0435\u0442 \u043F\u043E\
-  \u0440\u043E\u0433\u043E\u0432\u0443\u044E \u0434\u0430\u0442\u0443 \u0432 \u0437\
-  \u0430\u0432\u0438\u0441\u0438\u043C\u043E\u0441\u0442\u0438 \u043E\u0442 \u0442\
-  \u0435\u043A\u0443\u0449\u0435\u0433\u043E \u0434\u043D\u044F \u043C\u0435\u0441\
-  \u044F\u0446\u0430 \u0438 \u0444\u0438\u043B\u044C\u0442\u0440\u0443\u0435\u0442\
-  \ \u0434\u0430\u043D\u043D\u044B\u0435\r\n\r\nWITH date_threshold AS (\r\n    SELECT\
-  \  \r\n        -- \u041E\u043F\u0440\u0435\u0434\u0435\u043B\u044F\u0435\u043C \u043F\
-  \u043E\u0440\u043E\u0433\u043E\u0432\u0443\u044E \u0434\u0430\u0442\u0443 \u0432\
-  \ \u0437\u0430\u0432\u0438\u0441\u0438\u043C\u043E\u0441\u0442\u0438 \u043E\u0442\
-  \ \u0442\u0435\u043A\u0443\u0449\u0435\u0433\u043E \u0434\u043D\u044F  \r\n    \
-  \    CASE  \r\n            WHEN toDayOfMonth(now()) <= 10 THEN  \r\n           \
-  \     toStartOfMonth(dateSub(MONTH, 1, now()))  \r\n            ELSE  \r\n     \
-  \           toStartOfMonth(now())  \r\n        END AS cutoff_date  \r\n),\r\nfiltered_dates\
-  \ AS (\r\n    SELECT \r\n        dt,\r\n        formatDateTime(dt, '%d.%m.%Y') AS\
-  \ from_dt_txt,\r\n        formatDateTime(dt, '%d.%m.%Y') AS to_dt_txt\r\n      \
-  \ --dt as from_dt_txt,\r\n      -- dt as to_dt_txt\r\n    FROM dm_view.account_debt_for_working_capital_final\r\
-  \n    WHERE 1=1\r\n        -- \u0411\u0435\u0437\u043E\u043F\u0430\u0441\u043D\u0430\
-  \u044F \u043F\u0440\u043E\u0432\u0435\u0440\u043A\u0430 \u0444\u0438\u043B\u044C\
-  \u0442\u0440\u0430\r\n        {% if filter_values(\"yes_no_check\") | length !=\
-  \ 0 %}\r\n            {% if filter_values(\"yes_no_check\")[0] == \"\u0414\u0430\
-  \" %}\r\n                AND dt < (SELECT cutoff_date FROM date_threshold)\r\n \
-  \           {% endif %}\r\n        {% endif %}\r\n)\r\nSELECT \r\ndt,\r\n    from_dt_txt,\r\
-  \n    to_dt_txt,\r\n    formatDateTime(toLastDayOfMonth(dt), '%d.%m.%Y') as last_day_of_month_dt_txt\r\
-  \nFROM \r\n    filtered_dates\r\nGROUP BY \r\n    dt, from_dt_txt, to_dt_txt\r\n\
-  ORDER BY \r\n    dt DESC"
-params: null
-template_params: null
-filter_select_enabled: true
-fetch_values_predicate: null
-extra: null
-normalize_columns: false
-uuid: fca62707-6947-4440-a16b-70cb6a5cea5b
-metrics:
- metric_name: max_date
-  verbose_name: max_date
-  metric_type: count
-  expression: max(dt)
-  description: null
-  d3format: null
-  currency: null
-  extra:
-    warning_markdown: ''
-  warning_text: null
-columns:
- column_name: from_dt_txt
-  verbose_name: null
-  is_dttm: true
-  is_active: true
-  type: String
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: '%Y'
-  extra: {}
- column_name: dt
-  verbose_name: null
-  is_dttm: true
-  is_active: true
-  type: Date
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra: {}
- column_name: last_day_of_month_dt_txt
-  verbose_name: null
-  is_dttm: false
-  is_active: true
-  type: String
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra: {}
- column_name: to_dt_txt
-  verbose_name: null
-  is_dttm: true
-  is_active: true
-  type: String
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra: {}
-version: 1.0.0
-database_uuid: 97aced68-326a-4094-b381-27980560efa9
--- a/dashboards_example/dashboard_export_20260114T112108/datasets/Prod_Clickhouse_10/FI-0090_859.yaml
+++ b/dashboards_example/dashboard_export_20260114T112108/datasets/Prod_Clickhouse_10/FI-0090_859.yaml
@@ -1,190 +0,0 @@
-table_name: "FI-0090 \u0421\u0442\u0430\u0442\u0438\u0441\u0442\u0438\u043A\u0430\
-  \ \u043F\u043E \u0414\u0417/\u041F\u0414\u0417"
-main_dttm_col: dt
-description: null
-default_endpoint: null
-offset: 0
-cache_timeout: null
-schema: dm_view
-sql: "-- [JINJA_BLOCK] \u0426\u0435\u043D\u0442\u0440\u0430\u043B\u0438\u0437\u043E\
-  \u0432\u0430\u043D\u043D\u043E\u0435 \u043E\u043F\u0440\u0435\u0434\u0435\u043B\u0435\
-  \u043D\u0438\u0435 \u0432\u0441\u0435\u0445 Jinja \u043F\u0435\u0440\u0435\u043C\
-  \u0435\u043D\u043D\u044B\u0445\r\n{% set raw_to   = filter_values('last_day_of_month_dt_txt')[0]\
-  \ \r\n                  if filter_values('last_day_of_month_dt_txt')   else '01.05.2025'\
-  \ %}\r\n\r\n{# \u0440\u0430\u0437\u0431\u0438\u0432\u0430\u0435\u043C \xABDD.MM.YYYY\xBB\
-  \ \u043D\u0430 \u0447\u0430\u0441\u0442\u0438 #}\r\n{% set to_parts   = raw_to.split('.')\
-  \ %}\r\n\r\n{# \u0441\u043E\u0431\u0438\u0440\u0430\u0435\u043C ISO\u2011\u0441\u0442\
-  \u0440\u043E\u043A\u0443 \xABYYYY-MM-DD\xBB #}\r\n{% set to_dt   = to_parts[2] \
-  \  ~ '-' ~ to_parts[1]   ~ '-' ~ to_parts[0] %}\r\n\r\nwith \r\ncp_relations_type\
-  \ AS (\r\n   select * from ( SELECT \r\n        ctd.counterparty_code AS counterparty_code,\r\
-  \n        min(dt_from) as dt_from,\r\n        max(dt_to) as dt_to,\r\n        crt.relation_type_code\
-  \ || ' ' || crt.relation_type_name AS relation_type_code_name\r\n    FROM\r\n  \
-  \      dm_view.counterparty_td ctd\r\n    JOIN dm_view.counterparty_relation_type_texts\
-  \ crt \r\n        ON ctd.relation_type_code = crt.relation_type_code\r\n    GROUP\
-  \ BY\r\n        ctd.counterparty_code, ctd.counterparty_full_name,\r\n        crt.relation_type_code,crt.relation_type_name)\r\
-  \n            WHERE \r\n        dt_from <= toDate('{{to_dt }}') AND \r\n       \
-  \ dt_to >= toDate('{{to_dt }}')\r\n        ),\r\nt_debt as \r\n(SELECT dt, \r\n\
-  counterparty_search_name,\r\ncp_relations_type.relation_type_code_name as relation_type_code_name,\r\
-  \nunit_balance_code || ' ' || unit_balance_name as unit_balance_code_name,\r\n'1.\
-  \ \u0421\u0443\u043C\u043C\u0430' as attribute,\r\nsum(debt_balance_subposition_no_revaluation_usd_amount)\
-  \ as debt_amount,\r\nsumIf(debt_balance_subposition_no_revaluation_usd_amount,dt_overdue\
-  \ < dt) as overdue_amount\r\nfrom dm_view.account_debt_for_working_capital t_debt\r\
-  \njoin cp_relations_type ON\r\ncp_relations_type.counterparty_code = t_debt.counterparty_code\r\
-  \nwhere dt = toLastDayOfMonth(dt)\r\nand match(general_ledger_account_code,'((62)|(60)|(76))')\r\
-  \nand debit_or_credit = 'S'\r\nand account_type = 'D'\r\nand dt between addMonths(toDate('{{to_dt\
-  \ }}'),-12) and toDate('{{to_dt }}')\r\ngroup by dt, counterparty_search_name,unit_balance_code_name,relation_type_code_name\r\
-  \n),\r\n\r\nt_transaction_count_base as \r\n(\r\nselect *,\r\ncp_relations_type.relation_type_code_name\
-  \ as relation_type_code_name,\r\nunit_balance_code || ' ' || unit_balance_name as\
-  \ unit_balance_code_name,\r\n        case when dt_overdue<dt_clearing then\r\n \
-  \       dateDiff(day, dt_overdue, dt_clearing) \r\n        else 0\r\n        end\
-  \ as overdue_days\r\nfrom dm_view.accounting_documents_leading_to_debt t_docs\r\n\
-  join cp_relations_type ON\r\ncp_relations_type.counterparty_code = t_docs.counterparty_code\r\
-  \nwhere 1=1\r\n\r\nand match(general_ledger_account_code,'((62)|(60)|(76))')\r\n\
-  and debit_or_credit = 'S'\r\nand account_type = 'D'\r\n)\r\n\r\nselect * from t_debt\r\
-  \n\r\nunion all \r\n\r\nselect toLastDayOfMonth(dt_debt) as dt, \r\ncounterparty_search_name,\r\
-  \nrelation_type_code_name,\r\nunit_balance_code_name,\r\n'2. \u043A\u043E\u043B\u0438\
-  \u0447\u0435\u0441\u0442\u0432\u043E \u0442\u0440\u0430\u043D\u0437\u0430\u043A\u0446\
-  \u0438\u0439 \u0432 \u043C\u0435\u0441\u044F\u0446' as attribute,\r\ncount(1)  as\
-  \ debt_amount,\r\nnull as overdue_amount\r\nfrom t_transaction_count_base\r\nwhere\
-  \ dt_debt between addMonths(toDate('{{to_dt }}'),-12) and toDate('{{to_dt }}')\r\
-  \ngroup by toLastDayOfMonth(dt_debt), \r\ncounterparty_search_name,\r\nrelation_type_code_name,\r\
-  \nunit_balance_code_name,attribute\r\n\r\nunion all  \r\n\r\nselect toLastDayOfMonth(dt_clearing)\
-  \ as dt, \r\ncounterparty_search_name,\r\nrelation_type_code_name,\r\nunit_balance_code_name,\r\
-  \n'2. \u043A\u043E\u043B\u0438\u0447\u0435\u0441\u0442\u0432\u043E \u0442\u0440\u0430\
-  \u043D\u0437\u0430\u043A\u0446\u0438\u0439 \u0432 \u043C\u0435\u0441\u044F\u0446\
-  ' as attribute,\r\nnull  as debt_amount,\r\ncount(1) as overdue_amount\r\nfrom t_transaction_count_base\r\
-  \nwhere dt_clearing between addMonths(toDate('{{to_dt }}'),-12) and toDate('{{to_dt\
-  \ }}')\r\nand overdue_days > 0\r\ngroup by toLastDayOfMonth(dt_clearing), \r\ncounterparty_search_name,\r\
-  \nrelation_type_code_name,\r\nunit_balance_code_name,attribute\r\n\r\nunion all\
-  \ \r\n\r\nselect toLastDayOfMonth(dt_clearing) as dt, \r\ncounterparty_search_name,\r\
-  \nrelation_type_code_name,\r\nunit_balance_code_name,\r\nmultiIf(\r\noverdue_days\
-  \ < 30,'3. \u0434\u043E 30',\r\noverdue_days between 30 and 60, '4. \u043E\u0442\
-  \ 30 \u0434\u043E 60',\r\noverdue_days between 61 and 90, '5. \u043E\u0442 61 \u0434\
-  \u043E 90',\r\noverdue_days>90,'6. \u0431\u043E\u043B\u0435\u0435 90 \u0434\u043D\
-  ',\r\nnull\r\n)\r\n as attribute,\r\nnull  as debt_amount,\r\ncount(1) as overdue_amount\r\
-  \nfrom t_transaction_count_base\r\nwhere dt_clearing between addMonths(toDate('{{to_dt\
-  \ }}'),-12) and toDate('{{to_dt }}')\r\nand overdue_days > 0\r\ngroup by toLastDayOfMonth(dt_clearing),\
-  \ \r\ncounterparty_search_name,\r\nrelation_type_code_name,\r\nattribute,unit_balance_code_name,attribute\r\
-  \n"
-params: null
-template_params: null
-filter_select_enabled: true
-fetch_values_predicate: null
-extra: null
-normalize_columns: false
-uuid: 9e645dc0-da25-4f61-9465-6e649b0bc4b1
-metrics:
- metric_name: m_debt_amount
-  verbose_name: "\u0414\u0417, $"
-  metric_type: count
-  expression: sum(debt_amount)
-  description: null
-  d3format: null
-  currency: null
-  extra:
-    warning_markdown: ''
-  warning_text: null
- metric_name: m_overdue_amount
-  verbose_name: "\u041F\u0414\u0417, $"
-  metric_type: null
-  expression: sum(overdue_amount)
-  description: null
-  d3format: null
-  currency: null
-  extra:
-    warning_markdown: ''
-  warning_text: null
-columns:
- column_name: debt_amount
-  verbose_name: null
-  is_dttm: false
-  is_active: true
-  type: Nullable(Decimal(38, 2))
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra:
-    warning_markdown: null
- column_name: overdue_amount
-  verbose_name: null
-  is_dttm: false
-  is_active: true
-  type: Nullable(Decimal(38, 2))
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra:
-    warning_markdown: null
- column_name: dt
-  verbose_name: null
-  is_dttm: true
-  is_active: true
-  type: Nullable(Date)
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra:
-    warning_markdown: null
- column_name: unit_balance_code_name
-  verbose_name: null
-  is_dttm: false
-  is_active: true
-  type: Nullable(String)
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra:
-    warning_markdown: null
- column_name: relation_type_code_name
-  verbose_name: null
-  is_dttm: false
-  is_active: true
-  type: Nullable(String)
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra:
-    warning_markdown: null
- column_name: counterparty_search_name
-  verbose_name: null
-  is_dttm: false
-  is_active: true
-  type: Nullable(String)
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra:
-    warning_markdown: null
- column_name: attribute
-  verbose_name: null
-  is_dttm: false
-  is_active: true
-  type: Nullable(String)
-  advanced_data_type: null
-  groupby: true
-  filterable: true
-  expression: null
-  description: null
-  python_date_format: null
-  extra:
-    warning_markdown: null
-version: 1.0.0
-database_uuid: 97aced68-326a-4094-b381-27980560efa9
--- a/dashboards_example/dashboard_export_20260114T112108/metadata.yaml
+++ b/dashboards_example/dashboard_export_20260114T112108/metadata.yaml
@@ -1,3 +0,0 @@
-version: 1.0.0
-type: Dashboard
-timestamp: '2026-01-14T11:21:08.078620+00:00'
--- a/frontend/src/lib/api.js
+++ b/frontend/src/lib/api.js
@@ -100,6 +100,36 @@ async function fetchApi(endpoint) {
 }
 // [/DEF:fetchApi:Function]

+// [DEF:fetchApiBlob:Function]
+// @PURPOSE: Generic GET wrapper for binary payloads.
+// @PRE: endpoint string is provided.
+// @POST: Returns Blob or throws on error.
+async function fetchApiBlob(endpoint, options = {}) {
+    const notifyError = options.notifyError !== false;
+    try {
+        const response = await fetch(`${API_BASE_URL}${endpoint}`, {
+            headers: getAuthHeaders()
+        });
+        if (response.status === 202) {
+            const payload = await response.json().catch(() => ({ message: "Resource is being prepared" }));
+            const error = new Error(payload?.message || "Resource is being prepared");
+            error.status = 202;
+            throw error;
+        }
+        if (!response.ok) {
+            throw await buildApiError(response);
+        }
+        return await response.blob();
+    } catch (error) {
+        console.error(`[api.fetchApiBlob][Coherence:Failed] Error fetching blob from ${endpoint}:`, error);
+        if (notifyError) {
+            notifyApiError(error);
+        }
+        throw error;
+    }
+}
+// [/DEF:fetchApiBlob:Function]
+
 // [DEF:postApi:Function]
 // @PURPOSE: Generic POST request wrapper.
 // @PRE: endpoint and body are provided.
@@ -184,6 +214,16 @@ export const api = {
      return fetchApi(`/tasks${query ? `?${query}` : ''}`);
    },
    getTask: (taskId) => fetchApi(`/tasks/${taskId}`),
+    getTaskLogs: (taskId, options = {}) => {
+      const params = new URLSearchParams();
+      if (options.level) params.append('level', options.level);
+      if (options.source) params.append('source', options.source);
+      if (options.search) params.append('search', options.search);
+      if (options.offset != null) params.append('offset', String(options.offset));
+      if (options.limit != null) params.append('limit', String(options.limit));
+      const query = params.toString();
+      return fetchApi(`/tasks/${taskId}/logs${query ? `?${query}` : ''}`);
+    },
    createTask: (pluginId, params) => postApi('/tasks', { plugin_id: pluginId, params }),
    
    // Settings
@@ -198,6 +238,7 @@ export const api = {
    getStorageSettings: () => fetchApi('/settings/storage'),
    updateStorageSettings: (storage) => requestApi('/settings/storage', 'PUT', storage),
    getEnvironmentsList: () => fetchApi('/environments'),
+    getLlmStatus: () => fetchApi('/llm/status'),
    getEnvironmentDatabases: (id) => fetchApi(`/environments/${id}/databases`),
    
    // Dashboards
@@ -209,6 +250,18 @@ export const api = {
      return fetchApi(`/dashboards?${params.toString()}`);
    },
    getDashboardDetail: (envId, dashboardId) => fetchApi(`/dashboards/${dashboardId}?env_id=${envId}`),
+    getDashboardTaskHistory: (envId, dashboardId, options = {}) => {
+      const params = new URLSearchParams();
+      if (envId) params.append('env_id', envId);
+      if (options.limit) params.append('limit', options.limit);
+      return fetchApi(`/dashboards/${dashboardId}/tasks?${params.toString()}`);
+    },
+    getDashboardThumbnail: (envId, dashboardId, options = {}) => {
+      const params = new URLSearchParams();
+      params.append('env_id', envId);
+      if (options.force != null) params.append('force', String(Boolean(options.force)));
+      return fetchApiBlob(`/dashboards/${dashboardId}/thumbnail?${params.toString()}`, { notifyError: false });
+    },
    getDatabaseMappings: (sourceEnvId, targetEnvId) => fetchApi(`/dashboards/db-mappings?source_env_id=${sourceEnvId}&target_env_id=${targetEnvId}`),
    
     // Datasets
--- a/frontend/src/lib/components/assistant/AssistantChatPanel.svelte
+++ b/frontend/src/lib/components/assistant/AssistantChatPanel.svelte
@@ -21,6 +21,8 @@
   * @UX_TEST: LoadingHistory -> {openPanel: true, expected: loading block visible}
   * @UX_TEST: Sending -> {sendMessage: "branch", expected: send button disabled}
   * @UX_TEST: NeedsConfirmation -> {click: confirm action, expected: started response with task_id}
+   * @TEST_DATA: assistant_llm_ready -> {"llmStatus":{"configured":true,"reason":"ok"},"messages":[{"role":"assistant","text":"Ready","state":"success"}]}
+   * @TEST_DATA: assistant_llm_not_configured -> {"llmStatus":{"configured":false,"reason":"invalid_api_key"}}
   */

  import { onMount } from "svelte";
@@ -40,6 +42,7 @@
    getAssistantHistory,
    getAssistantConversations,
  } from "$lib/api/assistant.js";
+  import { api } from "$lib/api.js";
  import { gitService } from "../../../services/gitService.js";

  const HISTORY_PAGE_SIZE = 30;
@@ -62,6 +65,8 @@
  let conversationsHasNext = false;
  let historyViewport = null;
  let initialized = false;
+  let llmReady = true;
+  let llmStatusReason = "";

  $: isOpen = $assistantChatStore?.isOpen || false;
  $: conversationId = $assistantChatStore?.conversationId || null;
@@ -202,6 +207,7 @@
  $: if (isOpen && !initialized) {
    loadConversations(true);
    loadHistory();
+    loadLlmStatus();
  }

  $: if (isOpen && initialized && conversationId) {
@@ -502,6 +508,17 @@
  onMount(() => {
    initialized = false;
  });
+
+  async function loadLlmStatus() {
+    try {
+      const status = await api.getLlmStatus();
+      llmReady = Boolean(status?.configured);
+      llmStatusReason = status?.reason || "";
+    } catch (_err) {
+      llmReady = false;
+      llmStatusReason = "status_unavailable";
+    }
+  }
 </script>

 {#if isOpen}
@@ -533,6 +550,21 @@
    </div>

    <div class="flex h-[calc(100%-56px)] flex-col">
+      {#if !llmReady}
+        <div class="mx-3 mt-3 rounded-lg border border-rose-300 bg-rose-50 px-3 py-2 text-xs text-rose-800">
+          <div class="font-semibold">{$t.dashboard?.llm_not_configured || "LLM is not configured"}</div>
+          <div class="mt-1 text-rose-700">
+            {#if llmStatusReason === "no_active_provider"}
+              {$t.dashboard?.llm_configure_provider || "No active LLM provider. Configure it in Admin -> LLM Settings."}
+            {:else if llmStatusReason === "invalid_api_key"}
+              {$t.dashboard?.llm_configure_key || "Invalid LLM API key. Update and save a real key in Admin -> LLM Settings."}
+            {:else}
+              {$t.dashboard?.llm_status_unavailable || "LLM status is unavailable. Check settings and backend logs."}
+            {/if}
+          </div>
+        </div>
+      {/if}
+
      <div class="border-b border-slate-200 px-3 py-2">
        <div class="mb-2 flex items-center justify-between">
          <span
@@ -726,7 +758,7 @@
            bind:value={input}
            rows="2"
            placeholder={$t.assistant?.input_placeholder}
-            class="min-h-[52px] w-full resize-y rounded-lg border border-slate-300 px-3 py-2 text-sm outline-none transition focus:border-sky-400 focus:ring-2 focus:ring-sky-100"
+            class="min-h-[52px] w-full resize-y rounded-lg border px-3 py-2 text-sm outline-none transition {llmReady ? 'border-slate-300 focus:border-sky-400 focus:ring-2 focus:ring-sky-100' : 'border-rose-300 bg-rose-50 focus:border-rose-400 focus:ring-2 focus:ring-rose-100'}"
            on:keydown={handleKeydown}
          ></textarea>
          <button
--- a/frontend/src/lib/components/layout/TaskDrawer.svelte
+++ b/frontend/src/lib/components/layout/TaskDrawer.svelte
@@ -16,6 +16,8 @@
   * @UX_FEEDBACK: Back button returns to task list
   * @UX_RECOVERY: Click outside or X button closes drawer
   * @UX_RECOVERY: Back button shows task list when viewing task details
+   * @TEST_DATA: llm_task_success_with_fail_result -> {"activeTaskDetails":{"plugin_id":"llm_dashboard_validation","status":"SUCCESS","result":{"status":"FAIL"}}}
+   * @TEST_DATA: llm_task_success_with_pass_result -> {"activeTaskDetails":{"plugin_id":"llm_dashboard_validation","status":"SUCCESS","result":{"status":"PASS"}}}
   */

  import { onDestroy } from "svelte";
@@ -100,6 +102,22 @@
    );
  }

+  function resolveLlmValidationStatus(task) {
+    if (task?.plugin_id !== "llm_dashboard_validation") return null;
+    const raw = String(task?.result?.status || "").toUpperCase();
+    if (raw === "FAIL") return { label: "FAIL", tone: "fail", icon: "!" };
+    if (raw === "WARN") return { label: "WARN", tone: "warn", icon: "!" };
+    if (raw === "PASS") return { label: "PASS", tone: "pass", icon: "OK" };
+    return { label: "UNKNOWN", tone: "unknown", icon: "?" };
+  }
+
+  function llmValidationBadgeClass(tone) {
+    if (tone === "fail") return "text-rose-700 bg-rose-100 border border-rose-200";
+    if (tone === "warn") return "text-amber-700 bg-amber-100 border border-amber-200";
+    if (tone === "pass") return "text-emerald-700 bg-emerald-100 border border-emerald-200";
+    return "text-slate-700 bg-slate-100 border border-slate-200";
+  }
+
  function stopTaskDetailsPolling() {
    if (taskDetailsPollInterval) {
      clearInterval(taskDetailsPollInterval);
@@ -224,6 +242,18 @@
      return summary;
    }

+    if (task.plugin_id === "llm_dashboard_validation") {
+      summary.targetEnvId = resolveEnvironmentId(params?.environment_id || null);
+      summary.targetEnvName = resolveEnvironmentName(
+        summary.targetEnvId,
+        null,
+      );
+      if (result?.summary) {
+        summary.lines.push(result.summary);
+      }
+      return summary;
+    }
+
    if (result?.summary) {
      summary.lines.push(result.summary);
      return summary;
@@ -262,6 +292,15 @@
    }
  }

+  function handleOpenLlmReport() {
+    const taskId = normalizeTaskId(activeTaskId);
+    if (!taskId) {
+      addToast($t.tasks?.summary_link_unavailable || "Report unavailable", "error");
+      return;
+    }
+    window.open(`/reports/llm/${encodeURIComponent(taskId)}`, "_blank", "noopener,noreferrer");
+  }
+
  // Connect to WebSocket for real-time logs
  function connectWebSocket() {
    if (!activeTaskId) return;
@@ -401,6 +440,7 @@
  }

  $: taskSummary = buildTaskSummary(activeTaskDetails);
+  $: activeTaskValidation = resolveLlmValidationStatus(activeTaskDetails);

  // Cleanup on destroy
  onDestroy(() => {
@@ -461,6 +501,17 @@
            >{taskStatus}</span
          >
        {/if}
+        {#if activeTaskValidation}
+          <span
+            class={`text-xs font-semibold uppercase tracking-wider px-2 py-0.5 rounded-full inline-flex items-center gap-1 ${llmValidationBadgeClass(activeTaskValidation.tone)}`}
+            title="Dashboard validation result"
+          >
+            <span class="inline-flex min-w-[18px] items-center justify-center rounded-full bg-white/70 px-1 text-[10px] font-bold">
+              {activeTaskValidation.icon}
+            </span>
+            {activeTaskValidation.label}
+          </span>
+        {/if}
      </div>
      <div class="flex items-center gap-2">
        <button
@@ -536,6 +587,14 @@
              >
                {$t.tasks?.show_diff || "Show diff"}
              </button>
+              {#if activeTaskDetails?.plugin_id === "llm_dashboard_validation"}
+                <button
+                  class="rounded-md border border-indigo-300 bg-indigo-50 px-2.5 py-1.5 text-xs font-semibold text-indigo-700 transition-colors hover:bg-indigo-100"
+                  on:click={handleOpenLlmReport}
+                >
+                  {$t.tasks?.open_llm_report || "Open LLM report"}
+                </button>
+              {/if}
            </div>
            {#if showDiff}
              <div class="mt-3 rounded-md border border-slate-200 bg-white p-2">
@@ -576,6 +635,7 @@
            {$t.tasks?.recent }
          </h3>
          {#each recentTasks as task}
+            {@const taskValidation = resolveLlmValidationStatus(task)}
            <button
              class="flex items-center gap-3 w-full p-3 mb-2 bg-slate-800 border border-slate-700 rounded-lg cursor-pointer transition-all hover:bg-slate-700 hover:border-slate-600 text-left"
              on:click={() => selectTask(task)}
@@ -601,6 +661,17 @@
                      : 'bg-slate-500/15 text-slate-400'}"
                >{task.status || $t.common?.unknown }</span
              >
+              {#if taskValidation}
+                <span
+                  class={`text-[10px] font-semibold uppercase px-2 py-1 rounded-full inline-flex items-center gap-1 ${llmValidationBadgeClass(taskValidation.tone)}`}
+                  title="Dashboard validation result"
+                >
+                  <span class="inline-flex min-w-[16px] items-center justify-center rounded-full bg-white/70 px-1 text-[9px] font-bold">
+                    {taskValidation.icon}
+                  </span>
+                  {taskValidation.label}
+                </span>
+              {/if}
            </button>
          {/each}
        </div>
--- a/frontend/src/routes/dashboards/+page.svelte
+++ b/frontend/src/routes/dashboards/+page.svelte
@@ -5,7 +5,7 @@
   * @PURPOSE: Dashboard Hub - Central hub for managing dashboards with Git status and task actions
   * @LAYER: UI
   * @RELATION: BINDS_TO -> sidebarStore, taskDrawerStore
-   * @INVARIANT: Always shows environment selector and dashboard grid
+   * @INVARIANT: Always shows dashboards for the active environment from context store
   *
   * @UX_STATE: Loading -> Shows skeleton loader
   * @UX_STATE: Loaded -> Shows dashboard grid with status badges
@@ -29,7 +29,6 @@
  import {
    environmentContextStore,
    initializeEnvironmentContext,
-    setSelectedEnvironment,
  } from "$lib/stores/environmentContext.js";

  // State
@@ -160,15 +159,6 @@
    }
  }

-  // Handle environment change
-  function handleEnvChange(event) {
-    selectedEnv = event.target.value;
-    setSelectedEnvironment(selectedEnv);
-    currentPage = 1;
-    selectedIds.clear();
-    loadDashboards();
-  }
-
  // Handle search input
  function handleSearch(event) {
    debouncedSearch(event.target.value);
@@ -724,15 +714,6 @@
      {$t.nav?.dashboards }
    </h1>
    <div class="flex items-center space-x-4">
-      <select
-        class="px-4 py-2 border border-gray-300 rounded-lg bg-white focus:outline-none focus:ring-2 focus:ring-blue-500"
-        bind:value={selectedEnv}
-        on:change={handleEnvChange}
-      >
-        {#each environments as env}
-          <option value={env.id}>{env.name}</option>
-        {/each}
-      </select>
      <button
        class="inline-flex items-center justify-center rounded-lg bg-primary px-4 py-2 text-sm font-medium text-white transition-colors hover:bg-primary-hover"
        on:click={loadDashboards}
--- a/frontend/src/routes/dashboards/[id]/+page.svelte
+++ b/frontend/src/routes/dashboards/[id]/+page.svelte
@@ -6,13 +6,17 @@
   * @LAYER: UI
   * @RELATION: BINDS_TO -> dashboard detail API
   * @INVARIANT: Shows dashboard metadata, charts, and datasets for selected environment
+   * @TEST_DATA: dashboard_detail_ready -> {"dashboard":{"id":11,"title":"Ops","chart_count":3,"dataset_count":2},"taskHistory":[{"id":"t-1","plugin_id":"llm_dashboard_validation","status":"SUCCESS"}],"llmStatus":{"configured":true,"reason":"ok"}}
+   * @TEST_DATA: llm_unconfigured -> {"llmStatus":{"configured":false,"reason":"invalid_api_key"}}
   */

-  import { onMount } from "svelte";
+  import { onMount, onDestroy } from "svelte";
  import { goto } from "$app/navigation";
  import { page } from "$app/stores";
  import { t } from "$lib/i18n";
  import { api } from "$lib/api.js";
+  import { openDrawerForTask } from "$lib/stores/taskDrawer.js";
+  import { addToast } from "$lib/toasts.js";
  import Icon from "$lib/ui/Icon.svelte";

  $: dashboardId = $page.params.id;
@@ -21,11 +25,34 @@
  let dashboard = null;
  let isLoading = true;
  let error = null;
+  let taskHistory = [];
+  let isTaskHistoryLoading = false;
+  let taskHistoryError = null;
+  let isStartingBackup = false;
+  let isStartingValidation = false;
+  let thumbnailUrl = "";
+  let isThumbnailLoading = false;
+  let thumbnailError = null;
+  let llmReady = true;
+  let llmStatusReason = "";

  onMount(async () => {
-    await loadDashboardDetail();
+    await loadDashboardPage();
  });

+  onDestroy(() => {
+    releaseThumbnailUrl();
+  });
+
+  async function loadDashboardPage() {
+    await Promise.all([
+      loadDashboardDetail(),
+      loadTaskHistory(),
+      loadThumbnail(false),
+      loadLlmStatus(),
+    ]);
+  }
+
  async function loadDashboardDetail() {
    if (!dashboardId || !envId) {
      error = $t.dashboard?.missing_context ;
@@ -45,6 +72,144 @@
    }
  }

+  async function loadTaskHistory() {
+    if (!dashboardId || !envId) return;
+    isTaskHistoryLoading = true;
+    taskHistoryError = null;
+    try {
+      const response = await api.getDashboardTaskHistory(envId, dashboardId, { limit: 30 });
+      taskHistory = response?.items || [];
+    } catch (err) {
+      taskHistoryError = err.message || "Failed to load task history";
+      taskHistory = [];
+    } finally {
+      isTaskHistoryLoading = false;
+    }
+  }
+
+  function releaseThumbnailUrl() {
+    if (thumbnailUrl) {
+      URL.revokeObjectURL(thumbnailUrl);
+      thumbnailUrl = "";
+    }
+  }
+
+  async function loadThumbnail(force = false) {
+    if (!dashboardId || !envId) return;
+    isThumbnailLoading = true;
+    thumbnailError = null;
+    try {
+      const blob = await api.getDashboardThumbnail(envId, dashboardId, { force });
+      releaseThumbnailUrl();
+      thumbnailUrl = URL.createObjectURL(blob);
+    } catch (err) {
+      if (err?.status === 202) {
+        thumbnailError = $t.dashboard?.thumbnail_generating || "Thumbnail is being generated";
+      } else {
+        thumbnailError = err.message || $t.dashboard?.thumbnail_failed || "Failed to load thumbnail";
+      }
+    } finally {
+      isThumbnailLoading = false;
+    }
+  }
+
+  async function runBackupTask() {
+    if (isStartingBackup || !envId || !dashboardId) return;
+    isStartingBackup = true;
+    try {
+      const response = await api.postApi("/dashboards/backup", {
+        env_id: envId,
+        dashboard_ids: [Number(dashboardId)],
+      });
+      const taskId = response?.task_id;
+      if (taskId) {
+        openDrawerForTask(taskId);
+        addToast($t.dashboard?.backup_started || "Backup task started", "success");
+      }
+      await loadTaskHistory();
+    } catch (err) {
+      addToast(err.message || $t.dashboard?.backup_task_failed || "Failed to start backup", "error");
+    } finally {
+      isStartingBackup = false;
+    }
+  }
+
+  async function runLlmValidationTask() {
+    if (!llmReady) {
+      addToast($t.dashboard?.llm_not_configured || "LLM is not configured", "error");
+      return;
+    }
+    if (isStartingValidation || !envId || !dashboardId) return;
+    isStartingValidation = true;
+    try {
+      const response = await api.postApi("/tasks", {
+        plugin_id: "llm_dashboard_validation",
+        params: {
+          dashboard_id: String(dashboardId),
+          environment_id: envId,
+        },
+      });
+      const taskId = response?.task_id || response?.id;
+      if (taskId) {
+        openDrawerForTask(taskId);
+        addToast($t.dashboard?.validation_started || "LLM validation started", "success");
+      }
+      await Promise.all([
+        loadTaskHistory(),
+        loadThumbnail(true),
+      ]);
+    } catch (err) {
+      addToast(err.message || $t.dashboard?.validation_start_failed || "Failed to start LLM validation", "error");
+    } finally {
+      isStartingValidation = false;
+    }
+  }
+
+  function openLlmReport(taskId) {
+    if (!taskId) return;
+    window.open(`/reports/llm/${encodeURIComponent(String(taskId))}`, "_blank", "noopener,noreferrer");
+  }
+
+  function toTaskTypeLabel(pluginId) {
+    if (pluginId === "superset-backup") return $t.dashboard?.backup || "Backup";
+    if (pluginId === "llm_dashboard_validation") return $t.dashboard?.llm_check || "LLM Check";
+    return pluginId || "-";
+  }
+
+  function getTaskStatusClasses(status) {
+    const normalized = (status || "").toLowerCase();
+    if (normalized === "running" || normalized === "pending") return "bg-blue-100 text-blue-700";
+    if (normalized === "success") return "bg-emerald-100 text-emerald-700";
+    if (normalized === "failed" || normalized === "error") return "bg-rose-100 text-rose-700";
+    if (normalized === "awaiting_input" || normalized === "waiting_input") return "bg-amber-100 text-amber-700";
+    return "bg-slate-100 text-slate-700";
+  }
+
+  function getValidationStatus(task) {
+    if (task?.plugin_id !== "llm_dashboard_validation") {
+      return { label: "-", level: "na", icon: "" };
+    }
+    const rawStatus = String(task?.validation_status || "").toUpperCase();
+    if (rawStatus === "FAIL") {
+      return { label: "FAIL", level: "fail", icon: "!" };
+    }
+    if (rawStatus === "WARN") {
+      return { label: "WARN", level: "warn", icon: "!" };
+    }
+    if (rawStatus === "PASS") {
+      return { label: "PASS", level: "pass", icon: "OK" };
+    }
+    return { label: "UNKNOWN", level: "unknown", icon: "?" };
+  }
+
+  function getValidationStatusClasses(level) {
+    if (level === "fail") return "bg-rose-100 text-rose-700 border-rose-200";
+    if (level === "warn") return "bg-amber-100 text-amber-700 border-amber-200";
+    if (level === "pass") return "bg-emerald-100 text-emerald-700 border-emerald-200";
+    if (level === "unknown") return "bg-slate-100 text-slate-700 border-slate-200";
+    return "bg-slate-50 text-slate-400 border-slate-200";
+  }
+
  function goBack() {
    goto(`/dashboards?env_id=${encodeURIComponent(envId)}`);
  }
@@ -59,6 +224,17 @@
    if (Number.isNaN(parsed.getTime())) return "-";
    return `${parsed.toLocaleDateString()} ${parsed.toLocaleTimeString([], { hour: "2-digit", minute: "2-digit" })}`;
  }
+
+  async function loadLlmStatus() {
+    try {
+      const status = await api.getLlmStatus();
+      llmReady = Boolean(status?.configured);
+      llmStatusReason = status?.reason || "";
+    } catch (_err) {
+      llmReady = false;
+      llmStatusReason = "status_unavailable";
+    }
+  }
 </script>

 <div class="mx-auto w-full max-w-7xl space-y-6">
@@ -78,14 +254,46 @@
        {$t.common?.id }: {dashboardId}{#if dashboard?.slug} • {dashboard.slug}{/if}
      </p>
    </div>
-    <button
-      class="inline-flex items-center justify-center rounded-lg bg-primary px-4 py-2 text-sm font-medium text-white transition-colors hover:bg-primary-hover"
-      on:click={loadDashboardDetail}
-    >
-      {$t.common?.refresh }
-    </button>
+    <div class="flex flex-wrap items-center gap-2">
+      <button
+        class="inline-flex items-center justify-center rounded-lg border border-slate-300 bg-white px-4 py-2 text-sm font-medium text-slate-700 transition-colors hover:bg-slate-50"
+        on:click={runBackupTask}
+        disabled={isStartingBackup}
+      >
+        {isStartingBackup ? ($t.common?.loading || "Loading...") : ($t.dashboard?.run_backup || "Run backup")}
+      </button>
+      <button
+        class="inline-flex items-center justify-center rounded-lg border px-4 py-2 text-sm font-medium transition-colors {llmReady ? 'border-indigo-300 bg-indigo-50 text-indigo-700 hover:bg-indigo-100' : 'border-rose-300 bg-rose-50 text-rose-700 opacity-70 cursor-not-allowed'}"
+        on:click={runLlmValidationTask}
+        disabled={isStartingValidation || !llmReady}
+        title={!llmReady ? ($t.dashboard?.llm_not_configured || "LLM is not configured") : ""}
+      >
+        {isStartingValidation ? ($t.common?.loading || "Loading...") : ($t.dashboard?.run_llm_check || "Run LLM check")}
+      </button>
+      <button
+        class="inline-flex items-center justify-center rounded-lg bg-primary px-4 py-2 text-sm font-medium text-white transition-colors hover:bg-primary-hover"
+        on:click={loadDashboardPage}
+      >
+        {$t.common?.refresh }
+      </button>
+    </div>
  </div>

+  {#if !llmReady}
+    <div class="rounded-lg border border-rose-300 bg-rose-50 px-4 py-3 text-sm text-rose-800">
+      <div class="font-semibold">{$t.dashboard?.llm_not_configured || "LLM is not configured"}</div>
+      <div class="mt-1 text-rose-700">
+        {#if llmStatusReason === "no_active_provider"}
+          {$t.dashboard?.llm_configure_provider || "No active LLM provider. Configure it in Admin -> LLM Settings."}
+        {:else if llmStatusReason === "invalid_api_key"}
+          {$t.dashboard?.llm_configure_key || "Invalid LLM API key. Update and save a real key in Admin -> LLM Settings."}
+        {:else}
+          {$t.dashboard?.llm_status_unavailable || "LLM status is unavailable. Check settings and backend logs."}
+        {/if}
+      </div>
+    </div>
+  {/if}
+
  {#if error}
    <div class="flex items-center justify-between rounded-lg border border-red-300 bg-red-50 px-4 py-3 text-red-700">
      <span>{error}</span>
@@ -103,6 +311,115 @@
    </div>
    <div class="h-64 animate-pulse rounded-xl border border-slate-200 bg-white"></div>
  {:else if dashboard}
+    <div class="grid grid-cols-1 gap-6 xl:grid-cols-5">
+      <div class="rounded-xl border border-slate-200 bg-white p-4 xl:col-span-2">
+        <div class="mb-3 flex items-center justify-between">
+          <h2 class="text-sm font-semibold uppercase tracking-wide text-slate-500">
+            {$t.dashboard?.api_thumbnail || "Dashboard thumbnail"}
+          </h2>
+          <button
+            class="rounded-md border border-slate-300 px-2 py-1 text-xs text-slate-700 hover:bg-slate-50"
+            on:click={() => loadThumbnail(true)}
+            disabled={isThumbnailLoading}
+          >
+            {$t.common?.refresh || "Refresh"}
+          </button>
+        </div>
+        {#if isThumbnailLoading}
+          <div class="h-56 animate-pulse rounded-lg bg-slate-100"></div>
+        {:else if thumbnailUrl}
+          <img
+            src={thumbnailUrl}
+            alt="Dashboard thumbnail"
+            class="h-56 w-full rounded-lg border border-slate-200 object-cover"
+          />
+        {:else}
+          <div class="flex h-56 items-center justify-center rounded-lg border border-dashed border-slate-300 bg-slate-50 text-sm text-slate-500">
+            {thumbnailError || ($t.dashboard?.thumbnail_unavailable || "Thumbnail is unavailable")}
+          </div>
+        {/if}
+      </div>
+      <div class="rounded-xl border border-slate-200 bg-white p-4 xl:col-span-3">
+        <div class="mb-3 flex items-center justify-between">
+          <h2 class="text-sm font-semibold uppercase tracking-wide text-slate-500">
+            {$t.tasks?.recent || "Recent tasks"}
+          </h2>
+          <button
+            class="rounded-md border border-slate-300 px-2 py-1 text-xs text-slate-700 hover:bg-slate-50"
+            on:click={loadTaskHistory}
+            disabled={isTaskHistoryLoading}
+          >
+            {$t.common?.refresh || "Refresh"}
+          </button>
+        </div>
+        {#if isTaskHistoryLoading}
+          <div class="space-y-2">
+            {#each Array(4) as _}
+              <div class="h-10 animate-pulse rounded bg-slate-100"></div>
+            {/each}
+          </div>
+        {:else if taskHistoryError}
+          <div class="rounded-lg border border-rose-200 bg-rose-50 px-3 py-2 text-sm text-rose-700">{taskHistoryError}</div>
+        {:else if taskHistory.length === 0}
+          <div class="rounded-lg border border-slate-200 bg-slate-50 px-3 py-6 text-center text-sm text-slate-500">
+            {$t.tasks?.select_task || "No backup/LLM tasks yet"}
+          </div>
+        {:else}
+          <div class="overflow-x-auto">
+            <table class="min-w-full divide-y divide-slate-200 text-sm">
+              <thead class="bg-slate-50">
+                <tr>
+                  <th class="px-3 py-2 text-left font-semibold text-slate-600">{$t.common?.type || "Type"}</th>
+                  <th class="px-3 py-2 text-left font-semibold text-slate-600">{$t.common?.status || "Status"}</th>
+                  <th class="px-3 py-2 text-left font-semibold text-slate-600">{$t.tasks?.result || "Check"}</th>
+                  <th class="px-3 py-2 text-left font-semibold text-slate-600">{$t.common?.started || "Started"}</th>
+                  <th class="px-3 py-2 text-left font-semibold text-slate-600">{$t.common?.finished || "Finished"}</th>
+                  <th class="px-3 py-2 text-left font-semibold text-slate-600">{$t.common?.actions || "Actions"}</th>
+                </tr>
+              </thead>
+              <tbody class="divide-y divide-slate-100">
+                {#each taskHistory as task}
+                  {@const validation = getValidationStatus(task)}
+                  <tr>
+                    <td class="px-3 py-2 text-slate-800">{toTaskTypeLabel(task.plugin_id)}</td>
+                    <td class="px-3 py-2">
+                      <span class={`rounded-full px-2 py-1 text-xs font-semibold uppercase ${getTaskStatusClasses(task.status)}`}>
+                        {task.status}
+                      </span>
+                    </td>
+                    <td class="px-3 py-2">
+                      <span class={`inline-flex items-center gap-1 rounded-full border px-2 py-1 text-xs font-semibold uppercase ${getValidationStatusClasses(validation.level)}`}>
+                        {#if validation.icon}
+                          <span class="inline-flex min-w-[18px] items-center justify-center rounded-full bg-white/70 px-1 text-[10px] font-bold">
+                            {validation.icon}
+                          </span>
+                        {/if}
+                        {validation.label}
+                      </span>
+                    </td>
+                    <td class="px-3 py-2 text-slate-700">{formatDate(task.started_at)}</td>
+                    <td class="px-3 py-2 text-slate-700">{formatDate(task.finished_at)}</td>
+                    <td class="px-3 py-2">
+                      <div class="flex flex-wrap items-center gap-1">
+                        {#if task.plugin_id === "llm_dashboard_validation"}
+                          <button
+                            class="inline-flex items-center gap-1 rounded-md border border-indigo-300 bg-indigo-50 px-2 py-1 text-xs text-indigo-700 hover:bg-indigo-100"
+                            on:click={() => openLlmReport(task.id)}
+                          >
+                            {$t.tasks?.open_llm_report || "LLM report"}
+                          </button>
+                        {/if}
+                      </div>
+                    </td>
+                  </tr>
+                {/each}
+              </tbody>
+            </table>
+          </div>
+        {/if}
+      </div>
+    </div>
+
    <div class="grid grid-cols-1 gap-4 md:grid-cols-3">
      <div class="rounded-xl border border-slate-200 bg-white p-4">
        <p class="text-xs font-semibold uppercase tracking-wide text-slate-500">{$t.dashboard?.last_modified }</p>
--- a/frontend/src/routes/datasets/+page.svelte
+++ b/frontend/src/routes/datasets/+page.svelte
@@ -5,7 +5,7 @@
   * @PURPOSE: Dataset Hub - Dedicated hub for datasets with mapping progress
   * @LAYER: UI
   * @RELATION: BINDS_TO -> sidebarStore, taskDrawerStore
-   * @INVARIANT: Always shows environment selector and dataset grid
+   * @INVARIANT: Always shows datasets for the active environment from context store
   * 
   * @UX_STATE: Loading -> Shows skeleton loader
   * @UX_STATE: Loaded -> Shows dataset grid with mapping progress
--- a/frontend/src/routes/reports/llm/[taskId]/+page.svelte
+++ b/frontend/src/routes/reports/llm/[taskId]/+page.svelte
@@ -0,0 +1,260 @@
+<!-- [DEF:frontend/src/routes/reports/llm/[taskId]/+page.svelte:Component] -->
+<script>
+  /**
+   * @TIER: CRITICAL
+   * @PURPOSE: Full report page for LLM dashboard validation task execution.
+   * @LAYER: UI
+   * @RELATION: CALLS -> /api/tasks/{taskId}, /api/tasks/{taskId}/logs
+   * @RELATION: BINDS_TO -> page store
+   * @UX_STATE: Loading -> Skeleton placeholders visible.
+   * @UX_STATE: Loaded -> Full report with durations, screenshots, sent logs and task logs.
+   * @UX_STATE: Error -> Alert with retry action.
+   * @UX_FEEDBACK: Refresh button reloads report payload.
+   * @UX_RECOVERY: Retry via refresh button after API error.
+   * @TEST_DATA: llm_report_success -> {"task":{"id":"task-1","plugin_id":"llm_dashboard_validation","status":"SUCCESS","started_at":"2026-02-26T12:00:00Z","finished_at":"2026-02-26T12:00:10Z","result":{"summary":"OK","issues":[],"screenshot_paths":["/tmp/s1.png"],"logs_sent_to_llm":["[2026-02-26] explore_json"],"timings":{"validation_duration_ms":10000}}},"logs":[{"timestamp":"2026-02-26T12:00:01Z","level":"INFO","source":"llm","message":"Analyzing"}]}
+   * @TEST_DATA: llm_report_error -> {"error":"Failed to load LLM report"}
+   */
+  import { onMount } from "svelte";
+  import { page } from "$app/stores";
+  import { goto } from "$app/navigation";
+  import { api } from "$lib/api.js";
+  import { t } from "$lib/i18n";
+  import { openDrawerForTask } from "$lib/stores/taskDrawer.js";
+  import Icon from "$lib/ui/Icon.svelte";
+
+  $: taskId = $page.params.taskId;
+
+  let task = null;
+  let logs = [];
+  let isLoading = true;
+  let error = null;
+
+  function formatDate(value) {
+    if (!value) return "-";
+    const parsed = new Date(value);
+    if (Number.isNaN(parsed.getTime())) return "-";
+    return `${parsed.toLocaleDateString()} ${parsed.toLocaleTimeString([], { hour: "2-digit", minute: "2-digit", second: "2-digit" })}`;
+  }
+
+  function formatMs(value) {
+    const ms = Number(value);
+    if (!Number.isFinite(ms) || ms < 0) return "-";
+    if (ms < 1000) return `${ms} ms`;
+    return `${(ms / 1000).toFixed(2)} s`;
+  }
+
+  function getDashboardCheckResult(resultPayload) {
+    const rawStatus = String(resultPayload?.status || "").toUpperCase();
+    if (rawStatus === "FAIL") return { label: "FAIL", level: "fail", icon: "!" };
+    if (rawStatus === "WARN") return { label: "WARN", level: "warn", icon: "!" };
+    if (rawStatus === "PASS") return { label: "PASS", level: "pass", icon: "OK" };
+    return { label: "UNKNOWN", level: "unknown", icon: "?" };
+  }
+
+  function getCheckResultClasses(level) {
+    if (level === "fail") return "bg-rose-100 text-rose-700 border-rose-200";
+    if (level === "warn") return "bg-amber-100 text-amber-700 border-amber-200";
+    if (level === "pass") return "bg-emerald-100 text-emerald-700 border-emerald-200";
+    return "bg-slate-100 text-slate-700 border-slate-200";
+  }
+
+  async function loadReport() {
+    if (!taskId) return;
+    isLoading = true;
+    error = null;
+    try {
+      const [taskPayload, logsPayload] = await Promise.all([
+        api.getTask(taskId),
+        api.getTaskLogs(taskId, { limit: 1000 }),
+      ]);
+      task = taskPayload;
+      logs = Array.isArray(logsPayload) ? logsPayload : [];
+    } catch (err) {
+      error = err?.message || "Failed to load LLM report";
+    } finally {
+      isLoading = false;
+    }
+  }
+
+  function backToReports() {
+    goto("/reports");
+  }
+
+  function openTaskDetails() {
+    if (!taskId) return;
+    openDrawerForTask(taskId);
+  }
+
+  onMount(async () => {
+    await loadReport();
+  });
+
+  $: result = task?.result || {};
+  $: checkResult = getDashboardCheckResult(result);
+  $: timings = result?.timings || {};
+  $: screenshotPaths = Array.isArray(result?.screenshot_paths)
+    ? result.screenshot_paths
+    : (result?.screenshot_path ? [result.screenshot_path] : []);
+  $: sentLogs = Array.isArray(result?.logs_sent_to_llm) ? result.logs_sent_to_llm : [];
+</script>
+
+<div class="mx-auto w-full max-w-7xl space-y-6">
+  <div class="flex items-start justify-between">
+    <div>
+      <button
+        class="inline-flex items-center gap-2 rounded-lg px-2 py-1 text-sm text-slate-600 transition-colors hover:bg-slate-100 hover:text-slate-900"
+        on:click={backToReports}
+      >
+        <Icon name="chevronLeft" size={16} />
+        {$t.nav?.reports || "Reports"}
+      </button>
+      <h1 class="mt-2 text-2xl font-bold text-slate-900">
+        {$t.tasks?.result_llm_validation || "LLM Validation Report"}
+      </h1>
+      <p class="mt-1 text-sm text-slate-500">Task: {taskId}</p>
+    </div>
+    <div class="flex items-center gap-2">
+      <button
+        class="inline-flex items-center justify-center rounded-lg border border-slate-300 bg-white px-4 py-2 text-sm font-medium text-slate-700 transition-colors hover:bg-slate-50"
+        on:click={openTaskDetails}
+      >
+        {$t.tasks?.details_logs || "Task details and logs"}
+      </button>
+      <button
+        class="inline-flex items-center justify-center rounded-lg bg-primary px-4 py-2 text-sm font-medium text-white transition-colors hover:bg-primary-hover"
+        on:click={loadReport}
+      >
+        {$t.common?.refresh || "Refresh"}
+      </button>
+    </div>
+  </div>
+
+  {#if error}
+    <div class="rounded-lg border border-rose-300 bg-rose-50 px-4 py-3 text-rose-700">
+      <div>{error}</div>
+    </div>
+  {/if}
+
+  {#if isLoading}
+    <div class="space-y-3">
+      <div class="h-24 animate-pulse rounded-xl border border-slate-200 bg-white"></div>
+      <div class="h-64 animate-pulse rounded-xl border border-slate-200 bg-white"></div>
+      <div class="h-64 animate-pulse rounded-xl border border-slate-200 bg-white"></div>
+    </div>
+  {:else if task}
+    <div class="grid grid-cols-1 gap-4 md:grid-cols-5">
+      <div class="rounded-xl border border-slate-200 bg-white p-4">
+        <p class="text-xs font-semibold uppercase tracking-wide text-slate-500">Status</p>
+        <p class="mt-2 text-lg font-semibold text-slate-900">{task?.status || "-"}</p>
+      </div>
+      <div class="rounded-xl border border-slate-200 bg-white p-4">
+        <p class="text-xs font-semibold uppercase tracking-wide text-slate-500">Dashboard check</p>
+        <span class={`mt-2 inline-flex items-center gap-1 rounded-full border px-2 py-1 text-sm font-semibold uppercase ${getCheckResultClasses(checkResult.level)}`}>
+          <span class="inline-flex min-w-[18px] items-center justify-center rounded-full bg-white/70 px-1 text-[10px] font-bold">
+            {checkResult.icon}
+          </span>
+          {checkResult.label}
+        </span>
+      </div>
+      <div class="rounded-xl border border-slate-200 bg-white p-4">
+        <p class="text-xs font-semibold uppercase tracking-wide text-slate-500">Started</p>
+        <p class="mt-2 text-sm font-semibold text-slate-900">{formatDate(task?.started_at)}</p>
+      </div>
+      <div class="rounded-xl border border-slate-200 bg-white p-4">
+        <p class="text-xs font-semibold uppercase tracking-wide text-slate-500">Finished</p>
+        <p class="mt-2 text-sm font-semibold text-slate-900">{formatDate(task?.finished_at)}</p>
+      </div>
+      <div class="rounded-xl border border-slate-200 bg-white p-4">
+        <p class="text-xs font-semibold uppercase tracking-wide text-slate-500">Duration</p>
+        <p class="mt-2 text-lg font-semibold text-slate-900">{formatMs(timings?.validation_duration_ms)}</p>
+      </div>
+    </div>
+
+    <div class="rounded-xl border border-slate-200 bg-white p-4">
+      <h2 class="text-sm font-semibold uppercase tracking-wide text-slate-500">
+        Text Report
+      </h2>
+      <p class="mt-2 text-sm text-slate-800">{result?.summary || "-"}</p>
+      {#if Array.isArray(result?.issues) && result.issues.length > 0}
+        <div class="mt-4 space-y-2">
+          {#each result.issues as issue}
+            <div class="rounded-lg border border-slate-200 bg-slate-50 p-3 text-sm">
+              <div class="font-semibold text-slate-900">{issue?.severity || "INFO"}</div>
+              <div class="text-slate-700">{issue?.message || "-"}</div>
+              {#if issue?.location}
+                <div class="text-xs text-slate-500">Location: {issue.location}</div>
+              {/if}
+            </div>
+          {/each}
+        </div>
+      {/if}
+    </div>
+
+    <div class="rounded-xl border border-slate-200 bg-white p-4">
+      <h2 class="text-sm font-semibold uppercase tracking-wide text-slate-500">
+        Screenshots
+      </h2>
+      {#if screenshotPaths.length === 0}
+        <p class="mt-2 text-sm text-slate-500">No screenshots saved.</p>
+      {:else}
+        <div class="mt-3 grid grid-cols-1 gap-4 lg:grid-cols-2">
+          {#each screenshotPaths as path}
+            <a href={`/api/storage/file?path=${encodeURIComponent(path)}`} target="_blank" rel="noreferrer noopener" class="block">
+              <img
+                src={`/api/storage/file?path=${encodeURIComponent(path)}`}
+                alt="Validation screenshot"
+                class="h-64 w-full rounded-lg border border-slate-200 object-cover"
+              />
+              <p class="mt-1 truncate text-xs text-slate-500">{path}</p>
+            </a>
+          {/each}
+        </div>
+      {/if}
+    </div>
+
+    <div class="rounded-xl border border-slate-200 bg-white p-4">
+      <h2 class="text-sm font-semibold uppercase tracking-wide text-slate-500">
+        Logs sent to LLM ({sentLogs.length})
+      </h2>
+      {#if sentLogs.length === 0}
+        <p class="mt-2 text-sm text-slate-500">No source logs were attached.</p>
+      {:else}
+        <pre class="mt-3 max-h-80 overflow-auto rounded-lg bg-slate-900 p-3 text-xs text-slate-100">{sentLogs.join('\n')}</pre>
+      {/if}
+    </div>
+
+    <div class="rounded-xl border border-slate-200 bg-white p-4">
+      <h2 class="text-sm font-semibold uppercase tracking-wide text-slate-500">
+        Task execution logs ({logs.length})
+      </h2>
+      {#if logs.length === 0}
+        <p class="mt-2 text-sm text-slate-500">No task logs available.</p>
+      {:else}
+        <div class="mt-3 max-h-96 overflow-auto rounded-lg border border-slate-200">
+          <table class="min-w-full divide-y divide-slate-200 text-xs">
+            <thead class="bg-slate-50">
+              <tr>
+                <th class="px-3 py-2 text-left font-semibold text-slate-600">Time</th>
+                <th class="px-3 py-2 text-left font-semibold text-slate-600">Level</th>
+                <th class="px-3 py-2 text-left font-semibold text-slate-600">Source</th>
+                <th class="px-3 py-2 text-left font-semibold text-slate-600">Message</th>
+              </tr>
+            </thead>
+            <tbody class="divide-y divide-slate-100">
+              {#each logs as logEntry}
+                <tr>
+                  <td class="px-3 py-2 align-top text-slate-600">{formatDate(logEntry?.timestamp)}</td>
+                  <td class="px-3 py-2 align-top text-slate-700">{logEntry?.level || "-"}</td>
+                  <td class="px-3 py-2 align-top text-slate-700">{logEntry?.source || "-"}</td>
+                  <td class="px-3 py-2 align-top text-slate-800 whitespace-pre-wrap">{logEntry?.message || "-"}</td>
+                </tr>
+              {/each}
+            </tbody>
+          </table>
+        </div>
+      {/if}
+    </div>
+  {/if}
+</div>
+
+<!-- [/DEF:frontend/src/routes/reports/llm/[taskId]/+page.svelte:Component] -->
Author	SHA1	Message	Date
busya	4ff6d307f8	+ai update	2026-02-26 17:54:23 +03:00
busya	f4612c0737	Improve dashboard LLM validation UX and report flow	2026-02-26 17:53:41 +03:00