Fix release upload over SSH

Merge pull request 'fix: upload deployment archive to user home' (#44 ) from fix/cd-release-upload-path into main
Reviewed-on: #44
2026-09-14 18:13:51 +03:00 · 2026-09-14 14:33:14 +00:00 · 2026-09-14 17:23:29 +03:00 · 2026-09-14 13:58:41 +00:00 · 2026-09-14 15:28:02 +03:00 · 2026-09-14 15:24:41 +03:00
32 changed files with 965 additions and 2049 deletions
--- a/.env.example
+++ b/.env.example
@@ -10,6 +10,8 @@ CRAWL_LIMIT=
 REQUEST_TIMEOUT=30
 REQUEST_DELAY_SECONDS=1
 PARSER_USE_PLAYWRIGHT=false
+DISMISSAL_CONFIRMATION_RUNS=3
+MAX_AUTO_DISMISSALS_PER_RUN=25

 ADMIN_USERNAME=admin
 ADMIN_PASSWORD=change-me
--- a/.gitea/workflows/ci.yml
+++ b/.gitea/workflows/ci.yml
@@ -0,0 +1,89 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+  workflow_dispatch:
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    env:
+      DATABASE_URL: sqlite:///./ci.db
+      PYTHONUNBUFFERED: "1"
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+          cache: pip
+      - run: python -m pip install --upgrade pip
+      - run: python -m pip install -r requirements.txt
+      - run: python -m pytest -q
+      - run: node --check app/static/admin.js
+
+  deploy:
+    if: (github.event_name == 'push' && github.ref == 'refs/heads/main') || github.event_name == 'workflow_dispatch'
+    needs: test
+    runs-on: ubuntu-latest
+    concurrency:
+      group: production-deploy
+      cancel-in-progress: false
+    env:
+      PROD_PATH: /srv/miem_workers
+      RELEASE_NAME: miem-workers-release-${{ github.run_id }}.tar
+    steps:
+      - uses: actions/checkout@v4
+      - name: Prepare release
+        shell: bash
+        run: |
+          set -euo pipefail
+          tar --exclude=.git --exclude=.env -cf "$RUNNER_TEMP/$RELEASE_NAME" .
+          printf '%s\n' "$PROD_SSH_KEY" > "$RUNNER_TEMP/deploy_key"
+          printf '%s\n' "$PROD_KNOWN_HOSTS" > "$RUNNER_TEMP/known_hosts"
+          chmod 600 "$RUNNER_TEMP/deploy_key" "$RUNNER_TEMP/known_hosts"
+        env:
+          PROD_SSH_KEY: ${{ secrets.PROD_SSH_KEY }}
+          PROD_KNOWN_HOSTS: ${{ secrets.PROD_KNOWN_HOSTS }}
+      - name: Upload release
+        shell: bash
+        run: |
+          set -euo pipefail
+          ssh -i "$RUNNER_TEMP/deploy_key" \
+            -o BatchMode=yes \
+            -o UserKnownHostsFile="$RUNNER_TEMP/known_hosts" \
+            "$PROD_USER@$PROD_HOST" \
+            "cat > \"\$HOME/$RELEASE_NAME\"" \
+            < "$RUNNER_TEMP/$RELEASE_NAME"
+        env:
+          PROD_HOST: ${{ secrets.PROD_HOST }}
+          PROD_USER: ${{ secrets.PROD_USER }}
+      - name: Deploy and verify
+        shell: bash
+        run: |
+          set -euo pipefail
+          ssh -i "$RUNNER_TEMP/deploy_key" \
+            -o BatchMode=yes \
+            -o UserKnownHostsFile="$RUNNER_TEMP/known_hosts" \
+            "$PROD_USER@$PROD_HOST" \
+            "PROD_PATH='$PROD_PATH' RELEASE_NAME='$RELEASE_NAME' bash -s" <<'REMOTE'
+          set -euo pipefail
+          cd "$PROD_PATH"
+          release="$HOME/$RELEASE_NAME"
+          mkdir -p "$HOME/backups/miem_workers"
+          backup="$HOME/backups/miem_workers/miem_workers-$(date +%Y%m%d-%H%M%S).sql"
+          docker compose exec -T postgres sh -c 'pg_dump -U "$POSTGRES_USER" "$POSTGRES_DB"' > "$backup"
+          docker run --rm \
+            -v "$release:/src/release.tar:ro" \
+            -v "$PROD_PATH":/dst \
+            alpine:3.20 sh -c 'tar -xf /src/release.tar -C /dst'
+          docker compose up -d --build
+          curl --fail --silent --show-error http://127.0.0.1:8000/api/health
+          docker compose ps
+          rm -f "$release"
+          REMOTE
+        env:
+          PROD_HOST: ${{ secrets.PROD_HOST }}
+          PROD_USER: ${{ secrets.PROD_USER }}
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,5 +1,45 @@
 # Changelog

+## 0.8.7
+
+- Production CD передаёт release-архив через SSH вместо SFTP.
+
+## 0.8.6
+
+- Исправлена загрузка release-архива production CD в домашний каталог deploy-пользователя.
+
+## 0.8.5
+
+- Исправлена адаптивная компоновка фильтров справочника и скрытие skip-link до клавиатурного фокуса.
+
+## 0.8.4
+
+- Подключены к конфигурации ограничения массового автоматического увольнения, используемые плановым обходом.
+
+## 0.8.3
+
+- Удалены устаревшие пояснение о недоступном endpoint и проверка его отсутствия.
+
+## 0.8.2
+
+- Добавлен production CD workflow с backup PostgreSQL и проверкой healthcheck после обновления.
+
+## 0.8.1
+
+- Улучшены доступность, фильтры, состояния и адаптивность административного каталога.
+
+## 0.7.8
+
+- В колонке учёной степени отображаются все найденные степени без лишнего текста.
+
+## 0.7.7
+
+- Удалена неиспользуемая интеграция обмена данными и связанная документация и тесты.
+
+## 0.7.6
+
+- Возвращён еженедельный автоматический запуск обхода сотрудников.
+
 ## 0.7.5

 - Production Compose запускает только API и PostgreSQL.
--- a/MCP_DESCRIPTION.md
+++ b/MCP_DESCRIPTION.md
@@ -1,671 +0,0 @@
-# MCP: описание работы, структуры и тулзов
-
-Документ описывает MCP endpoint сервиса `miem-employees` по текущей реализации в `app/mcp.py`.
-
-## Где находится MCP
-
- FastAPI router: `app.mcp.router`
- Подключение к приложению: `app/main.py`
- HTTP endpoint: `POST /mcp`
- Локально при обычном запуске API: `http://localhost:8000/mcp`
- В Docker Compose endpoint обслуживает `api`: `http://localhost:8000/mcp`
- Авторизация на уровне приложения: отсутствует. Заголовок `Authorization` не проверяется и не влияет на ответ.
-
-Если доступ к MCP нужно ограничить, это должно делаться внешним контуром: bind на localhost, VPN, firewall, reverse proxy или отдельная сетевая политика.
-
-## Протокол
-
-Endpoint принимает JSON-RPC 2.0 over HTTP.
-
-Общий формат запроса:
-
-```json
-{
-  "jsonrpc": "2.0",
-  "id": 1,
-  "method": "tools/list",
-  "params": {}
-}
-```
-
-Общий формат успешного ответа:
-
-```json
-{
-  "jsonrpc": "2.0",
-  "id": 1,
-  "result": {}
-}
-```
-
-Общий формат ошибки:
-
-```json
-{
-  "jsonrpc": "2.0",
-  "id": 1,
-  "error": {
-    "code": -32601,
-    "message": "Method not found"
-  }
-}
-```
-
-Поддерживаемая версия MCP-протокола:
-
-```text
-2024-11-05
-```
-
-Имя сервиса:
-
-```text
-miem-employees
-```
-
-Версия сервера берется из `app.version.BACKEND_VERSION`.
-
-## Поддерживаемые JSON-RPC методы
-
-### initialize
-
-Возвращает метаданные MCP-сервера и capabilities.
-
-Запрос:
-
-```json
-{
-  "jsonrpc": "2.0",
-  "id": 1,
-  "method": "initialize",
-  "params": {}
-}
-```
-
-Ответ:
-
-```json
-{
-  "jsonrpc": "2.0",
-  "id": 1,
-  "result": {
-    "protocolVersion": "2024-11-05",
-    "serverInfo": {
-      "name": "miem-employees",
-      "version": "0.7.0"
-    },
-    "capabilities": {
-      "tools": {}
-    }
-  }
-}
-```
-
-### tools/list
-
-Возвращает список доступных tools с JSON Schema для аргументов.
-
-Запрос:
-
-```json
-{
-  "jsonrpc": "2.0",
-  "id": 1,
-  "method": "tools/list",
-  "params": {}
-}
-```
-
-Ответ содержит массив `result.tools`.
-
-### tools/call
-
-Вызывает один tool по имени.
-
-Запрос:
-
-```json
-{
-  "jsonrpc": "2.0",
-  "id": 1,
-  "method": "tools/call",
-  "params": {
-    "name": "search_employees",
-    "arguments": {
-      "query": "Сергеев",
-      "limit": 20
-    }
-  }
-}
-```
-
-Ответ tool всегда заворачивается в MCP content-массив:
-
-```json
-{
-  "jsonrpc": "2.0",
-  "id": 1,
-  "result": {
-    "content": [
-      {
-        "type": "text",
-        "text": "{\"items\":[]}"
-      }
-    ]
-  }
-}
-```
-
-Поле `text` содержит сериализованный JSON с `ensure_ascii=false`. Клиент должен распарсить это поле как JSON, если ему нужна структурированная нагрузка.
-
-## Ошибки
-
- Неизвестный JSON-RPC метод: `code = -32601`, `message = "Method not found"`.
- Исключения при обработке tool: `code = -32000`, `message` содержит текст исключения.
- Если сущность не найдена внутри отдельных tools, HTTP и JSON-RPC ответ остаются успешными, а полезная нагрузка содержит `{"error": "not_found"}`.
-
-## Источники данных
-
-MCP читает данные из основной базы через SQLAlchemy session из `app.db.get_db`.
-
-Основные таблицы и модели:
-
- `employees`: текущая карточка сотрудника, статус, профиль, `current_data`, checksum.
- `employee_publications`: нормализованные публикации сотрудников с авторами, DOI, аннотацией, описанием, citation text и raw JSON из HSE Publications.
- `employee_news_links`: нормализованные ссылки на новости из блока профиля «В новостях» с заголовком, URL, кратким описанием, датой, годом публикации и raw JSON карточки.
- `crawl_runs`: история запусков парсинга.
- `crawl_run_employee_changes`: детальные изменения сотрудников в рамках запуска.
- `crawl_errors`: ошибки парсинга в рамках запуска.
- `dataset_versions`: версии полного набора сотрудников.
- `dataset_version_items`: состав конкретной версии набора сотрудников.
-
-## Общая структура employee payload
-
-Краткая карточка сотрудника:
-
-```json
-{
-  "profile_key": "staff:avsergeev",
-  "profile_id": "avsergeev",
-  "full_name": "Сергеев Алексей Викторович",
-  "status": "active",
-  "canonical_url": "https://www.hse.ru/staff/avsergeev",
-  "last_seen_at": "2026-05-14T10:00:00+00:00",
-  "dismissed_at": null
-}
-```
-
-В sync payload дополнительно отдается `checksum`.
-
-Полная карточка дополнительно содержит:
-
-```json
-{
-  "data": {
-    "contacts": {},
-    "sections": []
-  }
-}
-```
-
-`data` соответствует распарсенному JSON профиля сотрудника. Внутри `sections` могут быть секции с публикациями, курсами, ВКР, новостями, таблицами, ссылками и произвольными текстовыми блоками.
-
-Пример секции новостей внутри `data.sections`:
-
-```json
-{
-  "title": "В новостях",
-  "slug": "v_novostyah",
-  "type": "news",
-  "news_count": 1,
-  "news_links": [
-    {
-      "title": "Название новости",
-      "url": "https://www.hse.ru/news/edu/1153850518.html",
-      "summary": "Краткое описание новости.",
-      "published_at": "2026-04-28T00:00:00+00:00",
-      "published_year": 2026
-    }
-  ]
-}
-```
-
-Для новостей отдельного MCP tool сейчас нет: они доступны через `get_employee(...).data.sections` или через полную синхронизацию `sync_employees(include_data=true)`.
-
-## Tools
-
-### get_service_info
-
-Назначение: вернуть метаданные сервиса, список tools и текущую версию набора сотрудников.
-
-Аргументы: отсутствуют.
-
-Возвращает:
-
-```json
-{
-  "service_name": "miem-employees",
-  "backend_version": "0.7.0",
-  "protocolVersion": "2024-11-05",
-  "tools": [],
-  "dataset": {
-    "hash": "sha256",
-    "previous_hash": "sha256 или null",
-    "created_at": "2026-05-14T10:00:00+00:00",
-    "crawl_run_id": 123,
-    "employee_count": 100,
-    "active_count": 95,
-    "dismissed_count": 5
-  }
-}
-```
-
-Особенность: перед ответом сервис создает актуальную `dataset_version`, если текущий набор сотрудников еще не имеет версии.
-
-### sync_employees
-
-Назначение: синхронизировать клиентский кэш сотрудников по hash набора данных.
-
-Аргументы:
-
-```json
-{
-  "client_hash": "sha256 или null",
-  "include_data": true
-}
-```
-
- `client_hash`: hash версии, которая уже есть у клиента. Если не передан, отдается полный snapshot.
- `include_data`: управляет включением полного `data` в карточки сотрудников. По умолчанию `true`.
-
-Полный ответ без `client_hash`:
-
-```json
-{
-  "mode": "full",
-  "from_hash": null,
-  "to_hash": "current-sha256",
-  "dataset": {},
-  "items": []
-}
-```
-
-Если клиентский hash совпадает с текущим:
-
-```json
-{
-  "mode": "delta",
-  "from_hash": "current-sha256",
-  "to_hash": "current-sha256",
-  "dataset": {},
-  "changes": {
-    "added": [],
-    "updated": [],
-    "dismissed": [],
-    "removed": []
-  }
-}
-```
-
-Если `client_hash` неизвестен серверу:
-
-```json
-{
-  "mode": "full",
-  "from_hash": "missing",
-  "to_hash": "current-sha256",
-  "dataset": {},
-  "items": [],
-  "reason": "unknown_client_hash"
-}
-```
-
-Если `client_hash` найден и отличается от текущего:
-
-```json
-{
-  "mode": "delta",
-  "from_hash": "old-sha256",
-  "to_hash": "current-sha256",
-  "dataset": {},
-  "changes": {
-    "added": [],
-    "updated": [],
-    "dismissed": [],
-    "removed": []
-  }
-}
-```
-
-Логика delta:
-
- `added`: сотрудник появился в новой версии.
- `updated`: изменился checksum или статус, и сотрудник активен.
- `dismissed`: сотрудник есть в новой версии, но получил статус `dismissed`.
- `removed`: `profile_key` был в старой версии, но отсутствует в новой.
-
-Hash набора считается по отсортированному списку `{profile_key, status, checksum}`.
-
-### search_employees
-
-Назначение: найти сотрудников по ФИО или canonical URL.
-
-Аргументы:
-
-```json
-{
-  "query": "Сергеев",
-  "status": "active",
-  "limit": 20
-}
-```
-
- `query`: обязательный по schema, но в коде пустая строка означает поиск без текстового фильтра.
- `status`: опционально, только `active` или `dismissed`.
- `limit`: максимум 100, по умолчанию 20.
-
-Возвращает массив кратких employee payload без `data`:
-
-```json
-[
-  {
-    "profile_key": "staff:avsergeev",
-    "profile_id": "avsergeev",
-    "full_name": "Сергеев Алексей Викторович",
-    "status": "active",
-    "canonical_url": "https://www.hse.ru/staff/avsergeev",
-    "last_seen_at": "2026-05-14T10:00:00+00:00",
-    "dismissed_at": null
-  }
-]
-```
-
-### get_employee
-
-Назначение: получить одну карточку сотрудника.
-
-Аргументы:
-
-```json
-{
-  "profile_id_or_url": "avsergeev"
-}
-```
-
-Поиск выполняется по:
-
- `profile_key`
- `profile_id`
- точному `canonical_url`
- частичному совпадению `canonical_url`
-
-Возвращает полный employee payload с `data`.
-
-Если сотрудник не найден:
-
-```json
-{
-  "error": "not_found"
-}
-```
-
-### list_employee_publications
-
-Назначение: вернуть публикации сотрудника. Если есть нормализованные строки в `employee_publications`, tool возвращает детальные публикационные данные: авторов, DOI, аннотацию, описание, citation text, год, тип, язык, статус и ссылки. Если детальная таблица еще не заполнена, tool использует старый fallback из `employees.current_data.sections[].publications`.
-
-Аргументы:
-
-```json
-{
-  "profile_id_or_url": "avsergeev"
-}
-```
-
-Поиск сотрудника выполняется так же, как в `get_employee`: по `profile_key`, `profile_id`, точному или частичному `canonical_url`.
-
-Порядок источников:
-
- сначала `employee_publications`, отсортированные по году, названию и внутреннему id;
- если записей нет, секции `current_data.sections` с `type = "publications"` и массивами `publications`.
-
-Ответ:
-
-```json
-{
-  "employee": {
-    "profile_key": "org_person:803294906",
-    "profile_id": "803294906",
-    "full_name": "Борисов Сергей Петрович",
-    "status": "active",
-    "canonical_url": "https://www.hse.ru/org/persons/803294906",
-    "last_seen_at": "2026-05-14T10:00:00+00:00",
-    "dismissed_at": null
-  },
-  "items": [
-    {
-      "id": "888959076",
-      "publication_id": "888959076",
-      "title": "Название публикации",
-      "text": "Краткое описание или citation",
-      "url": "https://publications.hse.ru/view/888959076",
-      "year": 2023,
-      "type": "ARTICLE",
-      "publication_type": "ARTICLE",
-      "language": "ru",
-      "status": 1,
-      "doi_url": "https://doi.org/10.53921/18195822_2023_23_4_624",
-      "other_url": "https://example.test",
-      "document_url": "https://example.test/file.pdf",
-      "citation_text": "Авторы. Название публикации // Журнал. 2023.",
-      "annotation": {
-        "ru": "Аннотация",
-        "en": "Abstract"
-      },
-      "description": {
-        "main": "Авторы. Название публикации // Журнал. 2023."
-      },
-      "authors": [
-        {
-          "id": "803294906",
-          "href": "https://www.hse.ru/org/persons/803294906",
-          "title_ru": "Борисов С. П.",
-          "title_en": "",
-          "reverse_title_ru": "С. П. Борисов",
-          "reverse_title_en": "",
-          "alt_name": "S. P. Borisov",
-          "other_name": null,
-          "is_current_employee": true
-        }
-      ]
-    }
-  ]
-}
-```
-
-В fallback-режиме из `current_data` старые элементы могут содержать только базовые поля `title`, `text`, `url` и `id`.
-
-Если сотрудник не найден:
-
-```json
-{
-  "items": []
-}
-```
-
-Если сотрудник найден, но публикаций нет:
-
-```json
-{
-  "employee": {},
-  "items": []
-}
-```
-
-### list_employee_courses
-
-Назначение: вернуть курсы преподавания сотрудника из распарсенных секций профиля.
-
-Аргументы:
-
-```json
-{
-  "profile_id_or_url": "avsergeev"
-}
-```
-
-Сервис ищет секции `current_data.sections` с `type = "courses_by_year"` и объединяет массивы `courses`.
-
-Ответ:
-
-```json
-{
-  "employee": {},
-  "items": [
-    {
-      "title": "Название курса",
-      "url": "https://..."
-    }
-  ]
-}
-```
-
-Если сотрудник или данные профиля отсутствуют:
-
-```json
-{
-  "items": []
-}
-```
-
-### get_crawl_status
-
-Назначение: вернуть последний запуск парсинга.
-
-Аргументы: отсутствуют.
-
-Ответ:
-
-```json
-{
-  "id": 123,
-  "status": "completed",
-  "source_url": "https://miem.hse.ru/persons",
-  "started_at": "2026-05-14T10:00:00+00:00",
-  "finished_at": "2026-05-14T10:10:00+00:00",
-  "found_count": 100,
-  "parsed_count": 98,
-  "error_count": 2,
-  "dismissed_count": 1
-}
-```
-
-Если запусков еще не было:
-
-```json
-{
-  "status": "never_run"
-}
-```
-
-### get_crawl_run_details
-
-Назначение: вернуть детальную информацию по конкретному запуску парсинга: summary, изменения сотрудников и ошибки.
-
-Аргументы:
-
-```json
-{
-  "run_id": 123
-}
-```
-
-Ответ:
-
-```json
-{
-  "id": 123,
-  "source_url": "https://miem.hse.ru/persons",
-  "status": "completed",
-  "status_display": "Завершен",
-  "started_at": "2026-05-14T10:00:00+00:00",
-  "finished_at": "2026-05-14T10:10:00+00:00",
-  "started_display": "14.05.2026 13:00",
-  "finished_display": "14.05.2026 13:10",
-  "found_count": 100,
-  "parsed_count": 98,
-  "new_count": 3,
-  "error_count": 2,
-  "dismissed_count": 1,
-  "processed_count": 100,
-  "progress_percent": 100.0,
-  "message": null,
-  "changes_detail_available": true,
-  "changes": {
-    "new": [],
-    "missing_from_source": [],
-    "dismissed": []
-  },
-  "errors": []
-}
-```
-
-Если запуск не найден:
-
-```json
-{
-  "error": "not_found"
-}
-```
-
-## Примеры curl
-
-Список tools:
-
-```bash
-curl http://localhost:8000/mcp \
-  -H "Content-Type: application/json" \
-  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}'
-```
-
-Поиск сотрудника:
-
-```bash
-curl http://localhost:8000/mcp \
-  -H "Content-Type: application/json" \
-  -d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"search_employees","arguments":{"query":"Сергеев","limit":5}}}'
-```
-
-Полная синхронизация:
-
-```bash
-curl http://localhost:8000/mcp \
-  -H "Content-Type: application/json" \
-  -d '{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"sync_employees","arguments":{"include_data":false}}}'
-```
-
-Delta-синхронизация:
-
-```bash
-curl http://localhost:8000/mcp \
-  -H "Content-Type: application/json" \
-  -d '{"jsonrpc":"2.0","id":4,"method":"tools/call","params":{"name":"sync_employees","arguments":{"client_hash":"known-sha256","include_data":true}}}'
-```
-
-## Как MCP используется клиентом
-
-1. Клиент вызывает `initialize` и проверяет `protocolVersion`.
-2. Клиент вызывает `tools/list`, чтобы получить актуальный список tools и input schemas.
-3. Для поиска и точечных запросов клиент вызывает `tools/call` с `search_employees`, `get_employee`, `list_employee_publications`, `list_employee_courses`, `get_crawl_status` или `get_crawl_run_details`.
-4. Для локального кэша клиент вызывает `get_service_info` или `sync_employees`.
-5. Клиент хранит последний `dataset.hash`.
-6. При следующей синхронизации клиент передает hash как `client_hash`.
-7. Сервер возвращает пустую delta, delta с изменениями или полный snapshot, если hash неизвестен.
-
-## Важные особенности реализации
-
- MCP endpoint read-only: tools не запускают парсинг и не меняют сотрудников напрямую.
- `get_service_info` и `sync_employees` могут создать новую запись `dataset_versions`, если состояние сотрудников изменилось и новой версии еще нет.
- Все tool payloads возвращаются как JSON-строка внутри `content[0].text`.
- `search_employees` ищет через `ilike` по `full_name` и `canonical_url`.
- `get_employee` допускает частичный URL, поэтому строка `133709486` может найти `https://www.hse.ru/org/persons/133709486`.
- Временные значения сериализуются через `isoformat()`, display-поля для админских payload формируются в часовом поясе `Europe/Moscow`.
--- a/README.md
+++ b/README.md
@@ -1,10 +1,11 @@
 # MIEM Employees Server

-Сервис собирает сотрудников МИЭМ с сайта ВШЭ, хранит карточки и историю обновлений в Postgres, показывает минимальную админку и отдает read-only MCP endpoint для ИИ-агентов.
+Сервис собирает сотрудников МИЭМ с сайта ВШЭ, хранит карточки и историю обновлений в Postgres и показывает минимальную админку.

 ## Архитектура

- `api`: FastAPI, REST API, HTML-админка, MCP endpoint и healthcheck.
+- `api`: FastAPI, REST API, HTML-админка и healthcheck.
+- `worker`: weekly scheduler, который запускает парсинг по `CRAWL_CRON`.
 - `postgres`: основная БД.

 Парсер использует фиксированный источник сотрудников, по умолчанию `https://miem.hse.ru/persons`. Для каждой карточки сохраняются ФИО, должности, год начала работы, контакты, идентификаторы, вкладки профиля, секции, публикации, курсы, ВКР, новости, JSON-снапшот и сжатый HTML-снапшот. Детальные публикации дополнительно нормализуются в отдельную таблицу `employee_publications`, а новости из блока «В новостях» — в `employee_news_links`. Ссылки обходятся только из меню профиля самого сотрудника (`person-menu`), например `#sci`, `#teaching`, `#main`.
@@ -20,7 +21,8 @@ cp .env.example .env
 Основные настройки:

 - `DATABASE_URL`: строка подключения SQLAlchemy.
- `SOURCE_URL`: список сотрудников МИЭМ.
+- `SOURCE_URL`: список сотрудников МИЭМ.
+- `CRAWL_CRON`: расписание в формате crontab, по умолчанию `0 3 * * 1`.
 - `CRAWL_LIMIT`: опциональный лимит профилей для тестового запуска.
 - `ADMIN_USERNAME`, `ADMIN_PASSWORD`: логин и пароль админки.
 - `SESSION_SECRET`: секрет подписи cookie.
@@ -41,9 +43,15 @@ uvicorn app.main:app --reload

 В админке доступны:

- `Dashboard`: общая статистика, последний добавленный сотрудник, прогресс текущего/последнего парсинга и ручной запуск.
- `Directory`: настраиваемая таблица сотрудников с фильтрами, сортировкой, пагинацией и выбором колонок.
- `Runs`: история запусков, ошибки и progress bar.
+- «Обзор»: статистика, последний добавленный сотрудник, прогресс парсинга и ручной запуск.
+- «Сотрудники»: поиск, фильтры, сортировка, пагинация и выбор колонок. По умолчанию показаны ФИО, статус, должности, дата последнего обнаружения и внешний профиль. В диалоге колонок доступны наборы «Проверка», «Контакты» и «Все поля»; ранее сохранённый выбор сохраняется.
+- «Запуски»: история обходов, ошибки и доступный индикатор прогресса.
+
+Все фильтры применяются кнопкой «Применить фильтры» с переходом на первую страницу. «Сбросить» очищает условия; при устаревшем номере страницы каталог возвращает первую. Подписи полей показывают выбранные условия, над таблицей указан диапазон результатов. Пояснения дат и статусов находятся под фильтрами.
+
+Каталог поддерживает навигацию с клавиатуры и системный диалог колонок с Escape и возвратом фокуса. На узких экранах таблица прокручивается внутри страницы, колонка ФИО закреплена. Используется системный шрифт без внешних загрузок.
+
+Каталог формируется сервером: отдельная клиентская загрузка списка не нужна, переход показывает браузер. Пустой результат предлагает сбросить фильтры, пустая база — запустить парсинг. При ошибке базы каталог возвращает HTTP 503 и предлагает повторить загрузку с теми же фильтрами. При обновлении прогресса показывается состояние загрузки; при ошибке сохраняются последние значения с предупреждением об их актуальности и кнопкой «Повторить». Автоматические попытки продолжаются каждые 4 секунды, запрос ограничен 15 секундами; параллельные запросы не запускаются.

 ## Docker Compose

@@ -54,7 +62,9 @@ docker compose up -d --build --remove-orphans
 По умолчанию:

 - API и админка: `http://localhost:8000`
- Postgres: `localhost:5432`
+- PostgreSQL: `postgres:5432` внутри сети Compose; порт на хост не опубликован.
+
+Compose запускает `api`, `worker` и `postgres`. API привязан к localhost; для внешнего доступа нужен настроенный reverse proxy. REST API данных требует сессию администратора.

 Таблицы создаются приложением при старте. При обновлении существующей базы приложение также добавляет недостающие runtime-колонки, например `crawl_runs.skipped_count`. SQL-миграции для ручного применения лежат в `migrations/`.

@@ -69,8 +79,6 @@ docker compose up -d --build --remove-orphans

 `employee_publications` содержит `publication_id`, название, год, тип публикации, язык, статус, ссылку на карточку HSE Publications, DOI, внешние/document-ссылки, citation text, аннотацию, описание, авторов, raw JSON ответа `searchPubs` и `source_hash` для безопасного повторного upsert. Уникальность поддерживается по `(employee_id, publication_id)` и `(employee_id, source_hash)`, поэтому повторный crawl не должен создавать дубликаты.

-`list_employee_publications` сначала читает `employee_publications`; если детальных строк еще нет, возвращает старые публикации из `current_data`.
-
 Новости сотрудников также хранятся в двух видах:

 - краткий список остается внутри `employees.current_data.sections[].news_links`;
@@ -80,7 +88,7 @@ docker compose up -d --build --remove-orphans

 ## Парсинг

-Ручной запуск доступен в админке на `Dashboard` и странице `Runs` или через REST:
+Worker запускает обход по `CRAWL_CRON`. Ручной запуск также доступен в админке на `Dashboard` и странице `Runs` или через REST:

 ```bash
 curl -X POST http://localhost:8000/api/crawl-runs --cookie "miem_admin_session=..."
@@ -105,41 +113,26 @@ curl -X POST http://localhost:8000/api/crawl-runs --cookie "miem_admin_session=.

 Во время выполнения парсинга `found_count`, `parsed_count`, `skipped_count` и `error_count` обновляются в базе. Админка опрашивает `/api/crawl-runs/latest` и показывает прогресс как `(parsed_count + skipped_count + error_count) / found_count`.

-## MCP
-
-Endpoint: `POST /mcp`, без авторизации на уровне приложения.
-
-Поддерживаемые tools:
-
- `get_service_info()`
- `sync_employees(client_hash?, include_data?)`
- `search_employees(query, status?, limit?)`
- `get_employee(profile_id_or_url)`
- `list_employee_publications(profile_id_or_url)` — публикации сотрудника; при наличии данных из `employee_publications` возвращает авторов, DOI, аннотацию, описание, citation text, год, тип, язык, статус и ссылку HSE Publications.
- `list_employee_courses(profile_id_or_url)`
- `get_crawl_status()`
- `get_crawl_run_details(run_id)`
-
-`get_service_info` возвращает метаданные сервиса, список tools и текущую версию набора сотрудников. `sync_employees` отдает полный snapshot или delta по `client_hash`; checksum набора строится по сотрудникам, их статусам и текущим checksums. Ответы tools возвращаются как JSON-строка внутри MCP `content[0].text`.
-
-Новости сотрудника отдельной MCP tool не имеют: они доступны в `get_employee(...).data.sections` и `sync_employees(include_data=true)` как секция `type = "news"` с массивом `news_links`.
-
-Пример локального запроса списка tools:
+## Обслуживание

 ```bash
-curl http://localhost:8000/mcp \
-  -H "Content-Type: application/json" \
-  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}'
-```
-
-Если MCP нужно ограничить, делайте это на сетевом уровне: localhost binding, VPN, firewall, reverse proxy или другой внешний контур доступа.
-
-## Обслуживание
-
-```bash
-docker compose logs -f api
+docker compose logs -f api
+docker compose logs -f worker
 docker compose exec postgres pg_dump -U miem miem_workers > backup.sql
-docker compose down
-```
-
-Версия сервиса: `0.7.5`. Админка всегда показывает версии backend и frontend в footer.
+docker compose down
+```
+
+Production deploy выполняется job `deploy` workflow `CI` после успешного `test` на push в `main` или вручную. Для него нужны Actions secrets `PROD_HOST`, `PROD_USER`, `PROD_SSH_KEY` и `PROD_KNOWN_HOSTS`; `.env` и ключи в репозиторий не добавляются. Перед обновлением workflow сохраняет PostgreSQL backup в домашний каталог deploy-пользователя, пересобирает `api` и `worker` и проверяет `/api/health`. Одновременно выполняется только один production deploy.
+
+## Проверки
+
+После установки зависимостей из `requirements.txt` выполните:
+
+```bash
+python -m pytest -q
+node --check app/static/admin.js
+```
+
+Тесты с данными используют временную SQLite; отдельные API smoke-тесты запускают приложение с его текущей конфигурацией. Браузерная проверка: `pip install playwright`, `python -m playwright install chromium`, затем `python tests/browser_admin.py`. Она проверяет реальный рендеринг страниц, клавиатуру, диалог, фильтры и восстановление прогресса после ошибки. Скриншоты сохраняются во временную папку; путь выводится в конце.
+
+Версия сервиса: `0.8.7`. Админка всегда показывает версии backend и frontend в footer.
--- a/app/admin.py
+++ b/app/admin.py
@@ -2,6 +2,7 @@ from fastapi import APIRouter, BackgroundTasks, Depends, Form, Request
 from fastapi.responses import HTMLResponse, RedirectResponse
 from fastapi.templating import Jinja2Templates
 from sqlalchemy import desc, func, select
+from sqlalchemy.exc import SQLAlchemyError
 from sqlalchemy.orm import Session

 from app.config import Settings, get_settings
@@ -93,19 +94,23 @@ def directory(
    parsed_started_to = _parse_date(started_to)
    parsed_has_email = None if has_email in (None, "") else has_email == "true"
    parsed_has_academic_degree = None if has_academic_degree in (None, "") else has_academic_degree == "true"
-    page = list_employees_page(
-        db,
-        status=status,
-        q=q,
-        started_from=parsed_started_from,
-        started_to=parsed_started_to,
-        has_email=parsed_has_email,
-        has_academic_degree=parsed_has_academic_degree,
-        sort=sort,
-        direction=direction,
-        limit=limit,
-        offset=offset,
-    )
+    try:
+        page = list_employees_page(
+            db,
+            status=status,
+            q=q,
+            started_from=parsed_started_from,
+            started_to=parsed_started_to,
+            has_email=parsed_has_email,
+            has_academic_degree=parsed_has_academic_degree,
+            sort=sort,
+            direction=direction,
+            limit=limit,
+            offset=offset,
+        )
+    except SQLAlchemyError:
+        db.rollback()
+        return _render(request, "directory_error.html", {}, status_code=503)
    return _render(
        request,
        "directory.html",
@@ -121,7 +126,7 @@ def directory(
                "sort": sort,
                "direction": direction,
                "limit": page["limit"],
-                "offset": offset,
+                "offset": page["offset"],
            },
        },
    )
--- a/app/config.py
+++ b/app/config.py
@@ -13,6 +13,8 @@ class Settings(BaseSettings):
    request_timeout: int = 30
    request_delay_seconds: float = 1.0
    parser_use_playwright: bool = False
+    dismissal_confirmation_runs: int = Field(default=3, ge=1)
+    max_auto_dismissals_per_run: int = Field(default=25, ge=1)

    admin_username: str = "admin"
    admin_password: str = "admin"
--- a/app/main.py
+++ b/app/main.py
@@ -4,14 +4,12 @@ from fastapi.staticfiles import StaticFiles
 from app.admin import router as admin_router
 from app.api import router as api_router
 from app.db import init_db
-from app.mcp import router as mcp_router
 from app.version import BACKEND_VERSION

 app = FastAPI(title="MIEM Employees", version=BACKEND_VERSION)
 app.mount("/static", StaticFiles(directory="app/static"), name="static")
 app.include_router(api_router)
 app.include_router(admin_router)
-app.include_router(mcp_router)


@app.on_event("startup")
--- a/app/mcp.py
+++ b/app/mcp.py
@@ -1,262 +0,0 @@
-import json
-
-from fastapi import APIRouter, Depends, Request
-from sqlalchemy import desc, or_, select
-from sqlalchemy.orm import Session
-
-from app.db import get_db
-from app.models import CrawlRun, Employee, EmployeePublication
-from app.services.admin_data import run_detail_payload
-from app.services.dataset_versions import service_info_payload, sync_employees_payload
-from app.version import BACKEND_VERSION
-
-router = APIRouter(prefix="/mcp")
-PROTOCOL_VERSION = "2024-11-05"
-SERVICE_NAME = "miem-employees"
-
-
-TOOLS = [
-    {
-        "name": "get_service_info",
-        "description": "Return service metadata, supported tools, and current dataset version.",
-        "inputSchema": {"type": "object", "properties": {}},
-    },
-    {
-        "name": "sync_employees",
-        "description": "Synchronize employees by dataset hash. Returns a full snapshot or a delta from client_hash.",
-        "inputSchema": {
-            "type": "object",
-            "properties": {
-                "client_hash": {"type": "string"},
-                "include_data": {"type": "boolean", "default": True},
-            },
-        },
-    },
-    {
-        "name": "search_employees",
-        "description": "Search MIEM employees by name or profile URL.",
-        "inputSchema": {
-            "type": "object",
-            "properties": {
-                "query": {"type": "string"},
-                "status": {"type": "string", "enum": ["active", "dismissed"]},
-                "limit": {"type": "integer", "default": 20},
-            },
-            "required": ["query"],
-        },
-    },
-    {
-        "name": "get_employee",
-        "description": "Get one employee by profile id, profile key, or canonical URL.",
-        "inputSchema": {"type": "object", "properties": {"profile_id_or_url": {"type": "string"}}, "required": ["profile_id_or_url"]},
-    },
-    {
-        "name": "list_employee_publications",
-        "description": (
-            "List employee publications with detailed fields when available: authors, DOI URL, annotation, "
-            "description, citation text, year, publication type, language, status, and HSE Publications URL."
-        ),
-        "inputSchema": {"type": "object", "properties": {"profile_id_or_url": {"type": "string"}}, "required": ["profile_id_or_url"]},
-    },
-    {
-        "name": "list_employee_courses",
-        "description": "List teaching courses parsed from an employee profile.",
-        "inputSchema": {"type": "object", "properties": {"profile_id_or_url": {"type": "string"}}, "required": ["profile_id_or_url"]},
-    },
-    {
-        "name": "get_crawl_status",
-        "description": "Return the latest crawl run status.",
-        "inputSchema": {"type": "object", "properties": {}},
-    },
-    {
-        "name": "get_crawl_run_details",
-        "description": "Return detailed employee changes and errors for one crawl run.",
-        "inputSchema": {
-            "type": "object",
-            "properties": {"run_id": {"type": "integer"}},
-            "required": ["run_id"],
-        },
-    },
-]
-
-
-@router.post("")
-async def mcp_http(
-    request: Request,
-    db: Session = Depends(get_db),
-) -> dict:
-    payload = await request.json()
-    method = payload.get("method")
-    request_id = payload.get("id")
-    params = payload.get("params") or {}
-
-    try:
-        if method == "initialize":
-            result = {
-                "protocolVersion": PROTOCOL_VERSION,
-                "serverInfo": {"name": SERVICE_NAME, "version": BACKEND_VERSION},
-                "capabilities": {"tools": {}},
-            }
-        elif method == "tools/list":
-            result = {"tools": TOOLS}
-        elif method == "tools/call":
-            result = _call_tool(db, params.get("name"), params.get("arguments") or {})
-        else:
-            return {"jsonrpc": "2.0", "id": request_id, "error": {"code": -32601, "message": "Method not found"}}
-        return {"jsonrpc": "2.0", "id": request_id, "result": result}
-    except Exception as exc:
-        return {"jsonrpc": "2.0", "id": request_id, "error": {"code": -32000, "message": str(exc)}}
-
-
-def _call_tool(db: Session, name: str, arguments: dict) -> dict:
-    if name == "get_service_info":
-        return _tool_response(
-            service_info_payload(
-                db,
-                tools=TOOLS,
-                service_name=SERVICE_NAME,
-                backend_version=BACKEND_VERSION,
-                protocol_version=PROTOCOL_VERSION,
-            )
-        )
-    if name == "sync_employees":
-        return _tool_response(
-            sync_employees_payload(
-                db,
-                client_hash=arguments.get("client_hash"),
-                include_data=bool(arguments.get("include_data", True)),
-            )
-        )
-    if name == "search_employees":
-        return _tool_response(_search_employees(db, arguments))
-    if name == "get_employee":
-        employee = _find_employee(db, arguments["profile_id_or_url"])
-        return _tool_response(_employee_payload(employee) if employee else {"error": "not_found"})
-    if name == "list_employee_publications":
-        employee = _find_employee(db, arguments["profile_id_or_url"])
-        return _tool_response(_collect_section_items(employee, "publications"))
-    if name == "list_employee_courses":
-        employee = _find_employee(db, arguments["profile_id_or_url"])
-        return _tool_response(_collect_section_items(employee, "courses_by_year"))
-    if name == "get_crawl_status":
-        run = db.scalar(select(CrawlRun).order_by(desc(CrawlRun.started_at)).limit(1))
-        return _tool_response(_run_payload(run) if run else {"status": "never_run"})
-    if name == "get_crawl_run_details":
-        run = db.get(CrawlRun, int(arguments["run_id"]))
-        return _tool_response(run_detail_payload(db, run) if run else {"error": "not_found"})
-    raise ValueError(f"Unknown tool: {name}")
-
-
-def _search_employees(db: Session, arguments: dict) -> list[dict]:
-    query = arguments.get("query", "")
-    limit = min(int(arguments.get("limit") or 20), 100)
-    stmt = select(Employee)
-    if arguments.get("status"):
-        stmt = stmt.where(Employee.status == arguments["status"])
-    if query:
-        pattern = f"%{query}%"
-        stmt = stmt.where(or_(Employee.full_name.ilike(pattern), Employee.canonical_url.ilike(pattern)))
-    employees = db.scalars(stmt.order_by(Employee.full_name).limit(limit)).all()
-    return [_employee_payload(employee, include_data=False) for employee in employees]
-
-
-def _find_employee(db: Session, value: str) -> Employee | None:
-    pattern = value.strip()
-    stmt = select(Employee).where(
-        or_(
-            Employee.profile_key == pattern,
-            Employee.profile_id == pattern,
-            Employee.canonical_url == pattern,
-            Employee.canonical_url.ilike(f"%{pattern}%"),
-        )
-    )
-    return db.scalar(stmt.limit(1))
-
-
-def _collect_section_items(employee: Employee | None, section_type: str) -> dict:
-    if not employee:
-        return {"items": []}
-    if section_type == "publications":
-        publications = _stored_publications(employee)
-        if publications:
-            return {"employee": _employee_payload(employee, include_data=False), "items": publications}
-    if not employee.current_data:
-        return {"employee": _employee_payload(employee, include_data=False), "items": []}
-    items = []
-    for section in employee.current_data.get("sections") or []:
-        if section.get("type") != section_type:
-            continue
-        if section_type == "publications":
-            items.extend(section.get("publications") or [])
-        elif section_type == "courses_by_year":
-            items.extend(section.get("courses") or [])
-    return {"employee": _employee_payload(employee, include_data=False), "items": items}
-
-
-def _stored_publications(employee: Employee) -> list[dict]:
-    return [_publication_payload(publication) for publication in sorted(employee.publications, key=_publication_sort_key)]
-
-
-def _publication_sort_key(publication: EmployeePublication) -> tuple:
-    return (publication.year or 0, publication.title or "", publication.id)
-
-
-def _publication_payload(publication: EmployeePublication) -> dict:
-    text = publication.citation_text or publication.title
-    payload = {
-        "id": publication.publication_id,
-        "publication_id": publication.publication_id,
-        "title": publication.title,
-        "text": text,
-        "url": publication.url,
-    }
-    optional = {
-        "year": publication.year,
-        "type": publication.publication_type,
-        "publication_type": publication.publication_type,
-        "language": publication.language,
-        "status": publication.status,
-        "doi_url": publication.doi_url,
-        "other_url": publication.other_url,
-        "document_url": publication.document_url,
-        "citation_text": publication.citation_text,
-        "annotation": publication.annotation,
-        "description": publication.description,
-        "authors": publication.authors,
-    }
-    payload.update({key: value for key, value in optional.items() if value not in (None, [], {})})
-    return payload
-
-
-def _employee_payload(employee: Employee, include_data: bool = True) -> dict:
-    payload = {
-        "profile_key": employee.profile_key,
-        "profile_id": employee.profile_id,
-        "full_name": employee.full_name,
-        "status": employee.status,
-        "canonical_url": employee.canonical_url,
-        "last_seen_at": employee.last_seen_at.isoformat() if employee.last_seen_at else None,
-        "dismissed_at": employee.dismissed_at.isoformat() if employee.dismissed_at else None,
-    }
-    if include_data:
-        payload["data"] = employee.current_data
-    return payload
-
-
-def _run_payload(run: CrawlRun) -> dict:
-    return {
-        "id": run.id,
-        "status": run.status,
-        "source_url": run.source_url,
-        "started_at": run.started_at.isoformat() if run.started_at else None,
-        "finished_at": run.finished_at.isoformat() if run.finished_at else None,
-        "found_count": run.found_count,
-        "parsed_count": run.parsed_count,
-        "skipped_count": run.skipped_count,
-        "error_count": run.error_count,
-        "dismissed_count": run.dismissed_count,
-    }
-
-
-def _tool_response(data: object) -> dict:
-    return {"content": [{"type": "text", "text": json.dumps(data, ensure_ascii=False, default=str)}]}
--- a/app/models.py
+++ b/app/models.py
@@ -162,7 +162,6 @@ class CrawlRun(Base):
    message: Mapped[str | None] = mapped_column(Text)

    employee_changes: Mapped[list["CrawlRunEmployeeChange"]] = relationship(back_populates="crawl_run")
-    dataset_versions: Mapped[list["DatasetVersion"]] = relationship(back_populates="crawl_run")


 class CrawlRunEmployeeChange(Base):
@@ -242,42 +241,3 @@ class ParseResourceCache(Base):
    body_snapshot: Mapped[bytes] = mapped_column(LargeBinary, nullable=False)
    parser_version: Mapped[str | None] = mapped_column(String(32))
    fetched_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), default=utcnow, nullable=False)
-
-
-class DatasetVersion(Base):
-    __tablename__ = "dataset_versions"
-    __table_args__ = (
-        UniqueConstraint("hash", name="uq_dataset_versions_hash"),
-        Index("ix_dataset_versions_created_at", "created_at"),
-    )
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True)
-    hash: Mapped[str] = mapped_column(String(64), nullable=False)
-    previous_hash: Mapped[str | None] = mapped_column(String(64))
-    crawl_run_id: Mapped[int | None] = mapped_column(ForeignKey("crawl_runs.id"))
-    employee_count: Mapped[int] = mapped_column(Integer, default=0, nullable=False)
-    active_count: Mapped[int] = mapped_column(Integer, default=0, nullable=False)
-    dismissed_count: Mapped[int] = mapped_column(Integer, default=0, nullable=False)
-    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), default=utcnow, nullable=False)
-
-    crawl_run: Mapped[CrawlRun | None] = relationship(back_populates="dataset_versions")
-    items: Mapped[list["DatasetVersionItem"]] = relationship(back_populates="dataset_version", cascade="all, delete-orphan")
-
-
-class DatasetVersionItem(Base):
-    __tablename__ = "dataset_version_items"
-    __table_args__ = (
-        UniqueConstraint("dataset_version_id", "profile_key", name="uq_dataset_version_items_version_profile"),
-        Index("ix_dataset_version_items_hash", "dataset_version_id"),
-        Index("ix_dataset_version_items_profile_key", "profile_key"),
-    )
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True)
-    dataset_version_id: Mapped[int] = mapped_column(ForeignKey("dataset_versions.id"), nullable=False)
-    profile_key: Mapped[str] = mapped_column(String(255), nullable=False)
-    employee_id: Mapped[int | None] = mapped_column(ForeignKey("employees.id"))
-    status: Mapped[str] = mapped_column(String(32), nullable=False)
-    checksum: Mapped[str] = mapped_column(String(64), nullable=False)
-
-    dataset_version: Mapped[DatasetVersion] = relationship(back_populates="items")
-    employee: Mapped[Employee | None] = relationship()
--- a/app/services/academic_degrees.py
+++ b/app/services/academic_degrees.py
@@ -2,13 +2,16 @@ import re
 from typing import Any


-_PATTERN = re.compile(r"\b(?:кандидат|доктор)\s+[\w\s-]{0,80}?\s+наук\b|\bph\.?\s*d\.?\b", re.IGNORECASE)
+_PATTERN = re.compile(r"\b(?:кандидат|доктор)(?:\s+[\w-]+){0,12}\s+наук\b|\bph\.?\s*d\.?(?!\w)", re.IGNORECASE)


 def academic_degrees(data: dict[str, Any] | None) -> list[str]:
    degrees = []
+    seen = set()
    for section in (data or {}).get("sections") or []:
-        if not isinstance(section, dict) or not re.search(r"уч[её]н.*степен|academic degree", str(section.get("title") or ""), re.IGNORECASE):
+        if not isinstance(section, dict) or not re.search(
+            r"образован|степен|academic degree|education", str(section.get("title") or ""), re.IGNORECASE
+        ):
            continue
        values = [
            *(entry.get("text") for entry in section.get("year_entries") or [] if isinstance(entry, dict)),
@@ -16,8 +19,15 @@ def academic_degrees(data: dict[str, Any] | None) -> list[str]:
            *(section.get("items") or []),
            section.get("raw_text"),
        ]
+        table = section.get("table") or {}
+        for row in table.get("rows") or []:
+            if isinstance(row, dict):
+                values.extend(row.get("cells") or [])
        for value in values:
            text = str(value or "").strip()
-            if text and _PATTERN.search(text) and text not in degrees:
-                degrees.append(text)
+            for match in _PATTERN.finditer(text):
+                degree = match.group(0).strip()
+                if degree.casefold() not in seen:
+                    seen.add(degree.casefold())
+                    degrees.append(degree)
    return degrees
--- a/app/services/admin_data.py
+++ b/app/services/admin_data.py
@@ -132,6 +132,8 @@ def list_employees_page(
        has_academic_degree=has_academic_degree,
    )
    total = db.scalar(select(func.count()).select_from(base_stmt.subquery())) or 0
+    if offset >= total:
+        offset = 0
    sort_column = EMPLOYEE_SORTS.get(sort, Employee.full_name)
    order = desc(sort_column) if direction == "desc" else sort_column
    employees = db.scalars(base_stmt.order_by(order).limit(limit).offset(offset)).all()
--- a/app/services/crawler.py
+++ b/app/services/crawler.py
@@ -26,7 +26,6 @@ from app.parser.collector import collect_profile_links
 from app.parser.profile import parse_person_profile
 from app.parser.profile_url import profile_key
 from app.services.academic_degrees import academic_degrees
-from app.services.dataset_versions import get_or_create_current_version
 from app.services.resource_cache import ResourceCache

 HEADERS = {
@@ -102,7 +101,6 @@ def run_crawl(db: Session, settings: Settings) -> CrawlRun:
                max_auto_dismissals=settings.max_auto_dismissals_per_run,
            )
        run.status = "completed"
-        get_or_create_current_version(db, crawl_run_id=run.id)
    except Exception as exc:
        run.status = "failed"
        run.message = str(exc)
@@ -146,7 +144,6 @@ def refresh_dismissed_status(db: Session, settings: Settings) -> CrawlRun:
            )
            run.parsed_count += 1
        run.status = "completed"
-        get_or_create_current_version(db, crawl_run_id=run.id)
    except Exception as exc:
        run.status = "failed"
        run.error_count = 1
@@ -194,7 +191,6 @@ def refresh_employee(db: Session, employee: Employee, settings: Settings) -> Cra
        else:
            run.skipped_count = 1
        run.status = "completed"
-        get_or_create_current_version(db, crawl_run_id=run.id)
    except Exception as exc:
        run.status = "failed"
        run.error_count = 1
--- a/app/services/dataset_versions.py
+++ b/app/services/dataset_versions.py
@@ -1,227 +0,0 @@
-import hashlib
-import json
-from dataclasses import dataclass
-
-from sqlalchemy import desc, select
-from sqlalchemy.orm import Session
-
-from app.models import DatasetVersion, DatasetVersionItem, Employee
-
-
-@dataclass(frozen=True)
-class EmployeeMarker:
-    profile_key: str
-    employee_id: int | None
-    status: str
-    checksum: str
-
-
-def get_or_create_current_version(db: Session, *, crawl_run_id: int | None = None) -> DatasetVersion:
-    employees = db.scalars(select(Employee).order_by(Employee.profile_key)).all()
-    markers = [_employee_marker(employee) for employee in employees]
-    dataset_hash = _dataset_hash(markers)
-    latest = get_latest_version(db)
-    if latest and latest.hash == dataset_hash:
-        return latest
-
-    active_count = sum(1 for marker in markers if marker.status == "active")
-    dismissed_count = sum(1 for marker in markers if marker.status == "dismissed")
-    version = DatasetVersion(
-        hash=dataset_hash,
-        previous_hash=latest.hash if latest else None,
-        crawl_run_id=crawl_run_id,
-        employee_count=len(markers),
-        active_count=active_count,
-        dismissed_count=dismissed_count,
-    )
-    db.add(version)
-    db.flush()
-    for marker in markers:
-        db.add(
-            DatasetVersionItem(
-                dataset_version_id=version.id,
-                profile_key=marker.profile_key,
-                employee_id=marker.employee_id,
-                status=marker.status,
-                checksum=marker.checksum,
-            )
-        )
-    db.flush()
-    return version
-
-
-def get_latest_version(db: Session) -> DatasetVersion | None:
-    return db.scalar(select(DatasetVersion).order_by(desc(DatasetVersion.created_at), desc(DatasetVersion.id)).limit(1))
-
-
-def get_version_by_hash(db: Session, dataset_hash: str | None) -> DatasetVersion | None:
-    if not dataset_hash:
-        return None
-    return db.scalar(select(DatasetVersion).where(DatasetVersion.hash == dataset_hash).limit(1))
-
-
-def service_info_payload(db: Session, *, tools: list[dict], service_name: str, backend_version: str, protocol_version: str) -> dict:
-    version = get_or_create_current_version(db)
-    db.commit()
-    return {
-        "service_name": service_name,
-        "backend_version": backend_version,
-        "protocolVersion": protocol_version,
-        "tools": tools,
-        "dataset": _version_payload(version),
-    }
-
-
-def sync_employees_payload(db: Session, *, client_hash: str | None = None, include_data: bool = True) -> dict:
-    current = get_or_create_current_version(db)
-    db.commit()
-    if not client_hash:
-        return _full_sync_payload(db, current, include_data=include_data, reason=None)
-    if client_hash == current.hash:
-        return {
-            "mode": "delta",
-            "from_hash": client_hash,
-            "to_hash": current.hash,
-            "dataset": _version_payload(current),
-            "changes": {"added": [], "updated": [], "dismissed": [], "removed": []},
-        }
-
-    previous = get_version_by_hash(db, client_hash)
-    if not previous:
-        return _full_sync_payload(db, current, include_data=include_data, reason="unknown_client_hash", from_hash=client_hash)
-
-    return _delta_sync_payload(db, previous, current, include_data=include_data)
-
-
-def _full_sync_payload(
-    db: Session,
-    current: DatasetVersion,
-    *,
-    include_data: bool,
-    reason: str | None,
-    from_hash: str | None = None,
-) -> dict:
-    employees = db.scalars(select(Employee).order_by(Employee.profile_key)).all()
-    payload = {
-        "mode": "full",
-        "from_hash": from_hash,
-        "to_hash": current.hash,
-        "dataset": _version_payload(current),
-        "items": [_employee_payload(employee, include_data=include_data) for employee in employees],
-    }
-    if reason:
-        payload["reason"] = reason
-    return payload
-
-
-def _delta_sync_payload(db: Session, previous: DatasetVersion, current: DatasetVersion, *, include_data: bool) -> dict:
-    previous_items = _items_by_profile_key(previous)
-    current_items = _items_by_profile_key(current)
-    employees = {employee.profile_key: employee for employee in db.scalars(select(Employee)).all()}
-    added = []
-    updated = []
-    dismissed = []
-    removed = []
-
-    for profile_key, current_item in sorted(current_items.items()):
-        previous_item = previous_items.get(profile_key)
-        employee = employees.get(profile_key)
-        if not previous_item:
-            if employee:
-                added.append(_employee_payload(employee, include_data=include_data))
-            continue
-        if previous_item.checksum == current_item.checksum and previous_item.status == current_item.status:
-            continue
-        if current_item.status == "dismissed":
-            dismissed.append(_tombstone(profile_key, current_item.status, employee))
-        elif employee:
-            updated.append(_employee_payload(employee, include_data=include_data))
-
-    for profile_key, previous_item in sorted(previous_items.items()):
-        if profile_key not in current_items:
-            removed.append(_tombstone(profile_key, "removed", employees.get(profile_key), checksum=previous_item.checksum))
-
-    return {
-        "mode": "delta",
-        "from_hash": previous.hash,
-        "to_hash": current.hash,
-        "dataset": _version_payload(current),
-        "changes": {
-            "added": added,
-            "updated": updated,
-            "dismissed": dismissed,
-            "removed": removed,
-        },
-    }
-
-
-def _items_by_profile_key(version: DatasetVersion) -> dict[str, DatasetVersionItem]:
-    return {item.profile_key: item for item in version.items}
-
-
-def _version_payload(version: DatasetVersion) -> dict:
-    return {
-        "hash": version.hash,
-        "previous_hash": version.previous_hash,
-        "created_at": version.created_at.isoformat() if version.created_at else None,
-        "crawl_run_id": version.crawl_run_id,
-        "employee_count": version.employee_count,
-        "active_count": version.active_count,
-        "dismissed_count": version.dismissed_count,
-    }
-
-
-def _employee_marker(employee: Employee) -> EmployeeMarker:
-    return EmployeeMarker(
-        profile_key=employee.profile_key,
-        employee_id=employee.id,
-        status=employee.status,
-        checksum=employee.current_checksum or _payload_hash(employee.current_data or {}),
-    )
-
-
-def _dataset_hash(markers: list[EmployeeMarker]) -> str:
-    payload = [
-        {"profile_key": marker.profile_key, "status": marker.status, "checksum": marker.checksum}
-        for marker in sorted(markers, key=lambda item: item.profile_key)
-    ]
-    return _payload_hash(payload)
-
-
-def _payload_hash(value: object) -> str:
-    payload = json.dumps(value, ensure_ascii=False, sort_keys=True, separators=(",", ":"), default=str)
-    return hashlib.sha256(payload.encode("utf-8")).hexdigest()
-
-
-def _employee_payload(employee: Employee, *, include_data: bool) -> dict:
-    payload = {
-        "profile_key": employee.profile_key,
-        "profile_id": employee.profile_id,
-        "full_name": employee.full_name,
-        "status": employee.status,
-        "canonical_url": employee.canonical_url,
-        "last_seen_at": employee.last_seen_at.isoformat() if employee.last_seen_at else None,
-        "dismissed_at": employee.dismissed_at.isoformat() if employee.dismissed_at else None,
-        "checksum": employee.current_checksum or _payload_hash(employee.current_data or {}),
-    }
-    if include_data:
-        payload["data"] = employee.current_data
-    return payload
-
-
-def _tombstone(profile_key: str, status: str, employee: Employee | None, *, checksum: str | None = None) -> dict:
-    payload = {
-        "profile_key": profile_key,
-        "status": status,
-        "checksum": checksum or (employee.current_checksum if employee else None),
-    }
-    if employee:
-        payload.update(
-            {
-                "profile_id": employee.profile_id,
-                "full_name": employee.full_name,
-                "canonical_url": employee.canonical_url,
-                "dismissed_at": employee.dismissed_at.isoformat() if employee.dismissed_at else None,
-            }
-        )
-    return payload
--- a/app/static/admin.css
+++ b/app/static/admin.css
@@ -5,9 +5,23 @@
  flex-direction: column;
  color: #1f2937;
  background: #f6f7f9;
-  font-family: Arial, sans-serif;
+  font-family: system-ui, -apple-system, BlinkMacSystemFont, "Segoe UI", sans-serif;
 }

+.admin__skip-link {
+  position: absolute;
+  top: 8px;
+  left: -999px;
+  z-index: 100;
+  padding: 8px 12px;
+  color: #ffffff;
+  background: #0f766e;
+}
+
+.admin__skip-link:focus-visible {
+  left: 8px;
+}
+
 .admin__header {
  display: flex;
  align-items: center;
@@ -16,6 +30,7 @@
  padding: 18px 32px;
  background: #ffffff;
  border-bottom: 1px solid #d9dee7;
+  flex-wrap: wrap;
 }

 .admin__brand {
@@ -40,6 +55,16 @@
  font-weight: 700;
 }

+.admin__link[aria-current="page"] {
+  text-decoration: underline;
+  text-underline-offset: 4px;
+}
+
+:is(a, button, input, select, summary):focus-visible {
+  outline: 3px solid #0f766e;
+  outline-offset: 2px;
+}
+
 .admin__main {
  flex: 1;
  width: min(1180px, calc(100% - 32px));
@@ -107,10 +132,6 @@
  border-collapse: collapse;
 }

-.table__row {
-  cursor: pointer;
-}
-
 .table__row:hover {
  background: #f0fdfa;
 }
@@ -161,8 +182,9 @@
  border-radius: 6px;
 }

-.button {
-  padding: 10px 14px;
+.button {
+  min-height: 40px;
+  padding: 10px 14px;
  border: 0;
  border-radius: 6px;
  color: #ffffff;
@@ -178,6 +200,19 @@

 .button--compact {
  padding: 8px 12px;
+}
+
+.button:hover {
+  filter: brightness(0.9);
+}
+
+.button:active {
+  filter: brightness(0.8);
+}
+
+.button:disabled {
+  cursor: wait;
+  opacity: 0.65;
 }

 .code {
@@ -426,25 +461,25 @@
  gap: 12px;
 }

-.progress-panel__header {
-  display: flex;
-  align-items: center;
-  justify-content: space-between;
-  gap: 16px;
-}
-
-.progress-panel__actions {
-  display: flex;
-  flex-wrap: wrap;
-  gap: 8px;
-}
-
-.button--secondary {
-  color: #0f766e;
-  background: #ccfbf1;
-}
-
-.progress-panel__body {
+.progress-panel__header {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 16px;
+}
+
+.progress-panel__actions {
+  display: flex;
+  flex-wrap: wrap;
+  gap: 8px;
+}
+
+.button--secondary {
+  color: #0f766e;
+  background: #ccfbf1;
+}
+
+.progress-panel__body {
  display: grid;
  gap: 10px;
 }
@@ -467,6 +502,12 @@
  color: #6b7280;
 }

+.progress-panel__error {
+  margin: 0;
+  color: #991b1b;
+  font-weight: 700;
+}
+
 .progress-bar {
  height: 12px;
  overflow: hidden;
@@ -478,7 +519,6 @@
  height: 100%;
  width: 0;
  background: #0f766e;
-  transition: width 0.25s ease;
 }

 .directory {
@@ -503,9 +543,9 @@
  color: #6b7280;
 }

-.directory__filters {
-  display: grid;
-  grid-template-columns: minmax(220px, 1.7fr) repeat(6, minmax(120px, 1fr));
+.directory__filters {
+  display: grid;
+  grid-template-columns: minmax(0, 1.1fr) minmax(0, 2fr) minmax(0, 1.25fr) minmax(0, 2fr);
  gap: 10px;
  padding: 16px;
  background: #ffffff;
@@ -513,13 +553,58 @@
  border-radius: 8px;
 }

-.directory__input {
+.directory__filter-group {
+  display: grid;
+  grid-template-columns: 1fr;
+  align-content: start;
+  gap: 8px;
+  min-width: 0;
+  margin: 0;
+  padding: 0;
+  border: 0;
+}
+
+.directory__filter-group--search {
+  grid-template-columns: 1fr;
+}
+
+.directory__filter-legend {
+  grid-column: 1 / -1;
+  margin-bottom: 2px;
+  color: #374151;
+  font-size: 12px;
+  font-weight: 700;
+}
+
+.directory__field {
+  display: grid;
+  gap: 5px;
+  min-width: 0;
+  color: #4b5563;
+  font-size: 12px;
+  font-weight: 700;
+}
+
+.directory__input {
+  box-sizing: border-box;
+  width: 100%;
+  font: inherit;
+  font-size: 16px;
+  font-weight: 400;
  min-width: 0;
  padding: 10px 12px;
  border: 1px solid #cbd5e1;
  border-radius: 6px;
 }

+.directory__filter-actions {
+  display: flex;
+  align-items: end;
+  flex-wrap: wrap;
+  gap: 8px;
+  grid-column: 1 / -1;
+}
+
 .directory__table-wrap {
  overflow-x: auto;
  background: #ffffff;
@@ -539,9 +624,18 @@
  font-weight: 700;
 }

+.directory__help {
+  color: #4b5563;
+  line-height: 1.5;
+}
+
+.directory__help summary {
+  cursor: pointer;
+}
+
 .directory-table {
  width: 100%;
-  min-width: 1120px;
+  min-width: 640px;
  border-collapse: collapse;
 }

@@ -556,14 +650,19 @@
 }

 .directory-table__cell {
+  overflow-wrap: anywhere;
  max-width: 280px;
  padding: 12px 10px;
  border-bottom: 1px solid #e5e7eb;
  vertical-align: top;
+}
+
+.directory-table .badge {
+  white-space: nowrap;
 }

 .directory-table__row {
-  cursor: pointer;
+  cursor: default;
 }

 .directory-table__row:hover {
@@ -576,33 +675,32 @@
  text-align: center;
 }

+.directory-table__empty span,
+.directory-table__empty a {
+  display: block;
+  margin-top: 8px;
+}
+
 .directory-table__cell--hidden,
 .directory-table__head--hidden {
  display: none;
 }

 .columns-modal {
-  position: fixed;
-  inset: 0;
-  z-index: 50;
-  display: grid;
-  place-items: center;
-  padding: 20px;
+  width: min(620px, calc(100% - 40px));
+  max-height: min(720px, calc(100vh - 40px));
+  margin: auto;
+  padding: 0;
+  border: 0;
+  background: transparent;
 }

-.columns-modal[hidden] {
-  display: none;
-}
-
-.columns-modal__backdrop {
-  position: absolute;
-  inset: 0;
+.columns-modal::backdrop {
  background: rgba(17, 24, 39, 0.54);
 }

 .columns-modal__panel {
  position: relative;
-  width: min(620px, 100%);
  max-height: min(720px, calc(100vh - 40px));
  overflow: auto;
  padding: 20px;
@@ -640,16 +738,37 @@
  border-radius: 6px;
 }

-.columns-modal__checkbox {
-  width: 16px;
-  height: 16px;
-}
+.columns-modal__checkbox {
+  width: 16px;
+  height: 16px;
+}
+
+.columns-modal__presets {
+  display: flex;
+  align-items: center;
+  flex-wrap: wrap;
+  gap: 8px;
+  margin-top: 18px;
+  padding: 0;
+  border: 0;
+}
+
+.columns-modal__preset-label {
+  color: #4b5563;
+  font-size: 13px;
+  font-weight: 700;
+}

@media (max-width: 920px) {
  .directory__filters {
    grid-template-columns: 1fr 1fr;
  }

+  .directory__filter-group--search,
+  .directory__filter-actions {
+    grid-column: 1 / -1;
+  }
+
  .progress-panel__header,
  .directory__header,
  .employee-card__header {
@@ -659,7 +778,40 @@
 }

@media (max-width: 620px) {
+  .directory-table [data-column="full_name"] {
+    position: sticky;
+    left: 0;
+    z-index: 1;
+    min-width: 130px;
+    max-width: 160px;
+    background: #ffffff;
+  }
+  .admin__header {
+    align-items: flex-start;
+    padding: 16px;
+  }
+
+  .admin__nav {
+    width: 100%;
+    flex-wrap: wrap;
+  }
+
+  .admin__main {
+    width: min(100% - 24px, 1180px);
+    margin: 20px auto;
+  }
+
  .directory__filters {
    grid-template-columns: 1fr;
  }
+
+  .directory__filter-group,
+  .directory__filter-group--search,
+  .directory__filter-actions {
+    grid-column: 1;
+  }
+
+  .directory__field {
+    font-size: 13px;
+  }
 }
--- a/app/static/admin.js
+++ b/app/static/admin.js
@@ -3,14 +3,15 @@
    "full_name",
    "status",
    "positions",
-    "hse_start_year",
-    "email",
-    "academic_degree",
    "last_seen_at",
-    "dismissed_at",
    "profile",
  ];
  const storageKey = "miem.directory.columns";
+  const columnPresets = {
+    review: ["full_name", "status", "academic_degree", "last_seen_at", "profile"],
+    contacts: ["full_name", "status", "email", "phone", "address", "profile"],
+    full: ["full_name", "status", "positions", "hse_start_year", "email", "phone", "address", "academic_degree", "publications_count", "courses_count", "news_count", "first_seen_at", "last_seen_at", "dismissed_at", "profile"],
+  };

  function readColumns() {
    try {
@@ -40,18 +41,33 @@
    if (!document.querySelector("[data-directory-table]")) return;
    let columns = readColumns();
    const modal = document.querySelector("[data-columns-modal]");
+    const trigger = document.querySelector("[data-columns-open]");
    applyColumns(columns);

-    document.querySelectorAll("[data-columns-open]").forEach((button) => {
-      button.addEventListener("click", () => {
-        modal.hidden = false;
-      });
+    trigger.addEventListener("click", () => {
+      modal.showModal();
+      trigger.setAttribute("aria-expanded", "true");
+      modal.querySelector("input")?.focus();
    });
    document.querySelectorAll("[data-columns-close]").forEach((button) => {
      button.addEventListener("click", () => {
-        modal.hidden = true;
+        modal.close();
      });
    });
+    document.querySelectorAll("[data-columns-preset]").forEach((button) => {
+      button.addEventListener("click", () => {
+        columns = columnPresets[button.dataset.columnsPreset];
+        writeColumns(columns);
+        applyColumns(columns);
+      });
+    });
+    modal.addEventListener("click", (event) => {
+      if (event.target === modal) modal.close();
+    });
+    modal.addEventListener("close", () => {
+      trigger.setAttribute("aria-expanded", "false");
+      trigger.focus();
+    });
    document.querySelectorAll("[data-column-toggle]").forEach((checkbox) => {
      checkbox.addEventListener("change", () => {
        columns = Array.from(document.querySelectorAll("[data-column-toggle]:checked")).map((item) => item.value);
@@ -62,28 +78,14 @@
    });
  }

-  function setupClickableRows() {
-    const openRow = (row) => {
-      window.location.href = row.dataset.rowHref;
-    };
-
-    document.querySelectorAll("[data-row-href]").forEach((row) => {
-      row.addEventListener("click", (event) => {
-        if (event.target.closest("a, button, input, select, label")) return;
-        openRow(row);
-      });
-      row.addEventListener("keydown", (event) => {
-        if (!["Enter", " "].includes(event.key)) return;
-        if (event.target.closest("a, button, input, select, label")) return;
-        event.preventDefault();
-        openRow(row);
-      });
-    });
-  }
-
  function setupProgress() {
    const panel = document.querySelector("[data-progress-panel]");
    if (!panel) return;
+    const state = document.createElement("p");
+    state.className = "progress-panel__empty";
+    state.setAttribute("role", "status");
+    panel.append(state);
+    let pending = false;

    const update = (run) => {
      if (!run) return;
@@ -94,6 +96,7 @@
      const errors = document.querySelector("[data-progress-errors]");
      const fill = document.querySelector("[data-progress-fill]");
      const percent = document.querySelector("[data-progress-percent]");
+      const error = document.querySelector("[data-progress-error]");
      if (status) status.textContent = run.status_display || run.status;
      if (processed) processed.textContent = run.processed_count;
      if (found) found.textContent = run.found_count;
@@ -101,21 +104,44 @@
      if (errors) errors.textContent = run.error_count;
      if (fill) fill.style.width = `${run.progress_percent}%`;
      if (percent) percent.textContent = run.progress_percent;
+      if (fill) {
+        fill.parentElement.setAttribute("aria-valuenow", run.progress_percent);
+        fill.parentElement.setAttribute("aria-valuetext", `${run.progress_percent}%`);
+      }
+      if (error) error.hidden = true;
    };

    const poll = async () => {
+      if (pending) return true;
+      pending = true;
+      state.textContent = "Обновляем прогресс…";
+      panel.setAttribute("aria-busy", "true");
      try {
-        const response = await fetch("/api/crawl-runs/latest", { credentials: "same-origin" });
-        if (!response.ok) return false;
+        const response = await fetch("/api/crawl-runs/latest", {
+          credentials: "same-origin",
+          signal: AbortSignal.timeout(15000),
+        });
+        if (!response.ok) throw new Error("progress request failed");
        const data = await response.json();
        const run = data.running || data.latest;
        update(run);
+        state.textContent = run ? "Прогресс обновлён" : "Запусков пока нет. Запустите парсинг.";
+        const error = document.querySelector("[data-progress-error]");
+        if (error) error.hidden = true;
        return Boolean(data.running);
      } catch (_error) {
-        return false;
+        const error = document.querySelector("[data-progress-error]");
+        if (error) error.hidden = false;
+        state.textContent = "Показаны последние полученные данные. Прогресс может быть устаревшим.";
+        return true;
+      } finally {
+        pending = false;
+        panel.setAttribute("aria-busy", "false");
      }
    };

+    document.querySelectorAll("[data-progress-retry]").forEach((button) => button.addEventListener("click", poll));
+
    const interval = window.setInterval(async () => {
      const keepGoing = await poll();
      if (!keepGoing) window.clearInterval(interval);
@@ -123,6 +149,5 @@
  }

  setupColumns();
-  setupClickableRows();
  setupProgress();
 })();
--- a/app/templates/base.html
+++ b/app/templates/base.html
@@ -7,18 +7,19 @@
    <link rel="stylesheet" href="/static/admin.css">
  </head>
  <body class="admin">
+    <a class="admin__skip-link" href="#main-content">Перейти к содержимому</a>
    <header class="admin__header">
      <h1 class="admin__brand"><a class="admin__brand-link" href="/admin">MIEM Employees</a></h1>
      <nav class="admin__nav">
-        <a class="admin__link" href="/admin">Обзор</a>
-        <a class="admin__link" href="/admin/directory">Сотрудники</a>
-        <a class="admin__link" href="/admin/runs">Запуски</a>
+        <a class="admin__link" href="/admin"{% if request.url.path == "/admin" %} aria-current="page"{% endif %}>Обзор</a>
+        <a class="admin__link" href="/admin/directory"{% if request.url.path == "/admin/directory" %} aria-current="page"{% endif %}>Сотрудники</a>
+        <a class="admin__link" href="/admin/runs"{% if request.url.path.startswith("/admin/runs") %} aria-current="page"{% endif %}>Запуски</a>
        <form method="post" action="/admin/logout">
          <button class="button button--ghost" type="submit">Выйти</button>
        </form>
      </nav>
    </header>
-    <main class="admin__main">
+    <main class="admin__main" id="main-content">
      {% block content %}{% endblock %}
    </main>
    <footer class="admin__footer">
--- a/app/templates/dashboard.html
+++ b/app/templates/dashboard.html
@@ -28,15 +28,15 @@
 </section>
 <section class="panel progress-panel" data-progress-panel>
  <div class="progress-panel__header">
-    <h2 class="panel__title">Прогресс парсинга</h2>
-    <div class="progress-panel__actions">
-      <form method="post" action="/admin/crawl-now">
-        <button class="button" type="submit">Запустить парсинг</button>
-      </form>
-      <form method="post" action="/admin/dismissed/refresh">
-        <button class="button button--secondary" type="submit">Проверить уволенных</button>
-      </form>
-    </div>
+    <h2 class="panel__title">Прогресс парсинга</h2>
+    <div class="progress-panel__actions">
+      <form method="post" action="/admin/crawl-now">
+        <button class="button" type="submit">Запустить парсинг</button>
+      </form>
+      <form method="post" action="/admin/dismissed/refresh">
+        <button class="button button--secondary" type="submit">Проверить уволенных</button>
+      </form>
+    </div>
  </div>
  {% set run = counts.current_running_run or latest_run %}
  <div class="progress-panel__body" data-progress-body>
@@ -46,10 +46,11 @@
      <span>без изменений: <span data-progress-skipped>{{ run.skipped_count if run else 0 }}</span></span>
      <span>ошибок: <span data-progress-errors>{{ run.error_count if run else 0 }}</span></span>
    </div>
-    <div class="progress-bar" aria-label="Parsing progress">
+    <div class="progress-bar" role="progressbar" aria-label="Прогресс парсинга" aria-valuemin="0" aria-valuemax="100" aria-valuenow="{{ run.progress_percent if run else 0 }}" aria-valuetext="{{ run.progress_percent if run else 0 }}%">
      <div class="progress-bar__fill" data-progress-fill style="width: {{ run.progress_percent if run else 0 }}%"></div>
    </div>
    <div class="progress-panel__percent"><span data-progress-percent>{{ run.progress_percent if run else 0 }}</span>%</div>
+    <p class="progress-panel__error" data-progress-error role="status" hidden>Не удалось обновить прогресс. <button class="button button--ghost button--compact" type="button" data-progress-retry>Повторить</button></p>
  </div>
 </section>
 <section class="panel">
@@ -58,7 +59,7 @@
    <thead><tr><th class="table__head">ID</th><th class="table__head">Статус</th><th class="table__head">Обработано</th><th class="table__head">Без изменений</th><th class="table__head">Ошибки</th><th class="table__head">Старт</th></tr></thead>
    <tbody>
      {% for run in runs %}
-      <tr class="table__row" onclick="window.location.href='/admin/runs/{{ run.id }}'" onkeydown="if (event.key === 'Enter' || event.key === ' ') { event.preventDefault(); window.location.href='/admin/runs/{{ run.id }}'; }" role="link" tabindex="0"><td class="table__cell">{{ run.id }}</td><td class="table__cell">{{ run.status_display }}</td><td class="table__cell">{{ run.parsed_count }}</td><td class="table__cell">{{ run.skipped_count }}</td><td class="table__cell">{{ run.error_count }}</td><td class="table__cell">{{ run.started_display }}</td></tr>
+      <tr class="table__row"><td class="table__cell"><a class="admin__link" href="/admin/runs/{{ run.id }}">{{ run.id }}</a></td><td class="table__cell">{{ run.status_display }}</td><td class="table__cell">{{ run.parsed_count }}</td><td class="table__cell">{{ run.skipped_count }}</td><td class="table__cell">{{ run.error_count }}</td><td class="table__cell">{{ run.started_display }}</td></tr>
      {% endfor %}
    </tbody>
  </table>
--- a/app/templates/directory.html
+++ b/app/templates/directory.html
@@ -5,49 +5,70 @@
  <div class="directory__header">
    <div>
      <h2 class="directory__title">Сотрудники</h2>
-      <p class="directory__summary">Найдено: {{ page.total }}</p>
+      {% set range_start = page.offset + 1 if page.total else 0 %}
+      {% set range_end = [page.offset + page.employees|length, page.total]|min %}
+      <p class="directory__summary" role="status">Показаны {{ range_start }}–{{ range_end }} из {{ page.total }}</p>
    </div>
-    <button class="button" type="button" data-columns-open>Колонки</button>
+    <button class="button" type="button" data-columns-open aria-controls="columns-dialog" aria-expanded="false">Настроить колонки</button>
  </div>

  <form class="directory__filters" method="get" action="/admin/directory">
-    <input class="directory__input" name="q" value="{{ filters.q }}" placeholder="ФИО или ссылка">
-    <select class="directory__input" name="status">
+    <fieldset class="directory__filter-group directory__filter-group--search">
+      <legend class="directory__filter-legend">Поиск</legend>
+      <label class="directory__field">ФИО или ссылка<input class="directory__input" name="q" value="{{ filters.q }}" placeholder="Например, Иванов или hse.ru"></label>
+    </fieldset>
+    <fieldset class="directory__filter-group">
+      <legend class="directory__filter-legend">Статус и данные</legend>
+      <label class="directory__field">Статус<select class="directory__input" name="status">
      <option value="" {% if not filters.status %}selected{% endif %}>Все статусы</option>
      <option value="active" {% if filters.status == "active" %}selected{% endif %}>Работает</option>
      <option value="verification_required" {% if filters.status == "verification_required" %}selected{% endif %}>Требует проверки</option>
      <option value="dismissed" {% if filters.status == "dismissed" %}selected{% endif %}>Уволен</option>
-    </select>
-    <select class="directory__input" name="has_email">
+      </select></label>
+      <label class="directory__field">Email<select class="directory__input" name="has_email">
      <option value="" {% if not filters.has_email %}selected{% endif %}>Любой email</option>
      <option value="true" {% if filters.has_email == "true" %}selected{% endif %}>Есть email</option>
      <option value="false" {% if filters.has_email == "false" %}selected{% endif %}>Нет email</option>
-    </select>
-    <select class="directory__input" name="has_academic_degree" aria-label="Учёная степень">
+      </select></label>
+      <label class="directory__field">Учёная степень<select class="directory__input" name="has_academic_degree">
      <option value="" {% if not filters.has_academic_degree %}selected{% endif %}>Любая учёная степень</option>
      <option value="true" {% if filters.has_academic_degree == "true" %}selected{% endif %}>Есть учёная степень</option>
      <option value="false" {% if filters.has_academic_degree == "false" %}selected{% endif %}>Нет учёной степени</option>
-    </select>
-    <input class="directory__input" type="date" name="started_from" value="{{ filters.started_from }}" aria-label="Впервые найден с">
-    <input class="directory__input" type="date" name="started_to" value="{{ filters.started_to }}" aria-label="Впервые найден по">
-    <select class="directory__input" name="sort">
+      </select></label>
+    </fieldset>
+    <fieldset class="directory__filter-group">
+      <legend class="directory__filter-legend">Период обнаружения</legend>
+      <label class="directory__field">От<input class="directory__input" type="date" name="started_from" value="{{ filters.started_from }}"></label>
+      <label class="directory__field">До<input class="directory__input" type="date" name="started_to" value="{{ filters.started_to }}"></label>
+    </fieldset>
+    <fieldset class="directory__filter-group">
+      <legend class="directory__filter-legend">Сортировка и страница</legend>
+      <label class="directory__field">Сортировать по<select class="directory__input" name="sort">
      {% for value, label in [("full_name", "ФИО"), ("status", "Статус"), ("hse_start_year", "Год начала"), ("first_seen_at", "Впервые найден"), ("last_seen_at", "Последний раз найден"), ("dismissed_at", "Дата увольнения")] %}
-      <option value="{{ value }}" {% if filters.sort == value %}selected{% endif %}>Сортировка: {{ label }}</option>
+      <option value="{{ value }}" {% if filters.sort == value %}selected{% endif %}>{{ label }}</option>
      {% endfor %}
-    </select>
-    <select class="directory__input" name="direction">
+      </select></label>
+      <label class="directory__field">Направление<select class="directory__input" name="direction">
      <option value="asc" {% if filters.direction == "asc" %}selected{% endif %}>По возрастанию</option>
      <option value="desc" {% if filters.direction == "desc" %}selected{% endif %}>По убыванию</option>
-    </select>
-    <select class="directory__input" name="limit" onchange="this.form.offset.value = 0; this.form.submit()">
+      </select></label>
+      <label class="directory__field">Сотрудников на странице<select class="directory__input" name="limit">
      {% for value in [25, 50, 100] %}
-      <option value="{{ value }}" {% if filters.limit == value %}selected{% endif %}>На странице: {{ value }}</option>
+      <option value="{{ value }}" {% if filters.limit == value %}selected{% endif %}>{{ value }}</option>
      {% endfor %}
-    </select>
-    <input type="hidden" name="offset" value="{{ filters.offset }}">
-    <button class="button" type="submit">Применить</button>
+      </select></label>
+    </fieldset>
+    <div class="directory__filter-actions">
+      <button class="button" type="submit">Применить фильтры</button>
+      <a class="button button--ghost" href="/admin/directory">Сбросить</a>
+    </div>
  </form>

+  <details class="directory__help">
+    <summary>Что означают даты и статусы</summary>
+    <p>«Год начала» — начало работы в ВШЭ. «Впервые найден» и «Последний раз найден» — даты обнаружения сотрудника в списке источника, а не даты приёма и увольнения. «Требует проверки» означает, что профиль недоступен и увольнение ещё не подтверждено.</p>
+  </details>
+
  <div class="directory__table-wrap">
    <table class="directory-table" data-directory-table>
      <thead>
@@ -71,9 +92,9 @@
      </thead>
      <tbody>
        {% for employee in page.employees %}
-        <tr class="directory-table__row" data-row-href="/admin/employees/{{ employee.id }}">
-          <td class="directory-table__cell" data-column="full_name">{{ employee.full_name or "Без имени" }}</td>
-          <td class="directory-table__cell" data-column="status"><span class="badge {% if employee.status == "dismissed" %}badge--dismissed{% endif %}">{{ employee.status_display }}</span></td>
+        <tr class="directory-table__row">
+          <td class="directory-table__cell" data-column="full_name"><a class="admin__link" href="/admin/employees/{{ employee.id }}">{{ employee.full_name or "Без имени" }}</a></td>
+          <td class="directory-table__cell" data-column="status"><span class="badge {% if employee.status == "dismissed" %}badge--dismissed{% elif employee.status == "verification_required" %}badge--verification{% endif %}">{{ employee.status_display }}</span></td>
          <td class="directory-table__cell" data-column="positions">{{ employee.positions_text }}</td>
          <td class="directory-table__cell" data-column="hse_start_year">{{ employee.hse_start_year or "" }}</td>
          <td class="directory-table__cell" data-column="email">{{ employee.email_text }}</td>
@@ -86,10 +107,18 @@
          <td class="directory-table__cell" data-column="first_seen_at">{{ employee.first_seen_display }}</td>
          <td class="directory-table__cell" data-column="last_seen_at">{{ employee.last_seen_display }}</td>
          <td class="directory-table__cell" data-column="dismissed_at">{{ employee.dismissed_display }}</td>
-          <td class="directory-table__cell" data-column="profile"><a class="admin__link" href="{{ employee.canonical_url }}">Открыть</a></td>
+          <td class="directory-table__cell" data-column="profile"><a class="admin__link" href="{{ employee.canonical_url }}" target="_blank" rel="noopener">Внешний профиль</a></td>
        </tr>
        {% else %}
-        <tr><td class="directory-table__empty" colspan="15">По этим фильтрам сотрудники не найдены.</td></tr>
+        <tr><td class="directory-table__empty" colspan="15">
+          <strong>Сотрудники не найдены</strong>
+          {% if filters.q or filters.status or filters.has_email or filters.has_academic_degree or filters.started_from or filters.started_to %}
+          <span>Попробуйте изменить запрос или сбросить фильтры.</span>
+          <a class="admin__link" href="/admin/directory">Сбросить фильтры</a>
+          {% else %}
+          <span>Запустите парсинг, чтобы загрузить сотрудников.</span>
+          {% endif %}
+        </td></tr>
        {% endfor %}
      </tbody>
    </table>
@@ -108,11 +137,10 @@
  </div>
 </section>

-<div class="columns-modal" data-columns-modal hidden>
-  <div class="columns-modal__backdrop" data-columns-close></div>
-  <section class="columns-modal__panel" aria-label="Настройка колонок">
+<dialog class="columns-modal" id="columns-dialog" data-columns-modal aria-labelledby="columns-dialog-title">
+  <section class="columns-modal__panel">
    <div class="columns-modal__header">
-      <h3 class="columns-modal__title">Отображаемые колонки</h3>
+      <h3 class="columns-modal__title" id="columns-dialog-title">Отображаемые колонки</h3>
      <button class="button button--ghost" type="button" data-columns-close>Закрыть</button>
    </div>
    <div class="columns-modal__grid">
@@ -120,8 +148,14 @@
      <label class="columns-modal__option"><input class="columns-modal__checkbox" type="checkbox" value="{{ key }}" data-column-toggle> {{ label }}</label>
      {% endfor %}
    </div>
+    <fieldset class="columns-modal__presets">
+      <legend class="columns-modal__preset-label">Быстрый набор</legend>
+      <button class="button button--ghost button--compact" type="button" data-columns-preset="review">Проверка</button>
+      <button class="button button--ghost button--compact" type="button" data-columns-preset="contacts">Контакты</button>
+      <button class="button button--ghost button--compact" type="button" data-columns-preset="full">Все поля</button>
+    </fieldset>
  </section>
-</div>
+</dialog>
 {% endblock %}
 {% block scripts %}
 <script src="/static/admin.js"></script>
--- a/app/templates/directory_error.html
+++ b/app/templates/directory_error.html
@@ -0,0 +1,9 @@
+{% extends "base.html" %}
+{% block title %}Каталог недоступен · MIEM Employees{% endblock %}
+{% block content %}
+<section class="panel">
+  <h2 class="panel__title">Не удалось загрузить сотрудников</h2>
+  <p>База данных временно недоступна. Повторите запрос; выбранные фильтры сохранятся.</p>
+  <a class="button" href="{{ request.url.path }}{% if request.url.query %}?{{ request.url.query }}{% endif %}">Повторить загрузку</a>
+</section>
+{% endblock %}
--- a/app/templates/runs.html
+++ b/app/templates/runs.html
@@ -17,10 +17,11 @@
      <span>без изменений: <span data-progress-skipped>{{ run.skipped_count }}</span></span>
      <span>ошибок: <span data-progress-errors>{{ run.error_count }}</span></span>
    </div>
-    <div class="progress-bar" aria-label="Parsing progress">
+    <div class="progress-bar" role="progressbar" aria-label="Прогресс парсинга" aria-valuemin="0" aria-valuemax="100" aria-valuenow="{{ percent }}" aria-valuetext="{{ percent }}%">
      <div class="progress-bar__fill" data-progress-fill style="width: {{ percent }}%"></div>
    </div>
  <div class="progress-panel__percent"><span data-progress-percent>{{ percent }}</span>%</div>
+  <p class="progress-panel__error" data-progress-error role="status" hidden>Не удалось обновить прогресс. <button class="button button--ghost button--compact" type="button" data-progress-retry>Повторить</button></p>
  </div>
  {% else %}
  <div class="progress-panel" data-progress-panel>
@@ -30,17 +31,18 @@
      <span>без изменений: <span data-progress-skipped>0</span></span>
      <span>ошибок: <span data-progress-errors>0</span></span>
    </div>
-    <div class="progress-bar" aria-label="Parsing progress">
+    <div class="progress-bar" role="progressbar" aria-label="Прогресс парсинга" aria-valuemin="0" aria-valuemax="100" aria-valuenow="0" aria-valuetext="0%">
      <div class="progress-bar__fill" data-progress-fill style="width: 0%"></div>
    </div>
    <div class="progress-panel__percent"><span data-progress-percent>0</span>%</div>
+    <p class="progress-panel__error" data-progress-error role="status" hidden>Не удалось обновить прогресс. <button class="button button--ghost button--compact" type="button" data-progress-retry>Повторить</button></p>
  </div>
  {% endif %}
  <table class="table">
    <thead><tr><th class="table__head">ID</th><th class="table__head">Статус</th><th class="table__head">Найдено</th><th class="table__head">Обработано</th><th class="table__head">Без изменений</th><th class="table__head">Новые</th><th class="table__head">Ошибки</th><th class="table__head">Уволены</th><th class="table__head">Старт</th></tr></thead>
    <tbody>
      {% for run in runs %}
-      <tr class="table__row" onclick="window.location.href='/admin/runs/{{ run.id }}'" onkeydown="if (event.key === 'Enter' || event.key === ' ') { event.preventDefault(); window.location.href='/admin/runs/{{ run.id }}'; }" role="link" tabindex="0"><td class="table__cell">{{ run.id }}</td><td class="table__cell">{{ run.status_display }}</td><td class="table__cell">{{ run.found_count }}</td><td class="table__cell">{{ run.parsed_count }}</td><td class="table__cell">{{ run.skipped_count }}</td><td class="table__cell">{{ run.new_count }}</td><td class="table__cell">{{ run.error_count }}</td><td class="table__cell">{{ run.dismissed_count }}</td><td class="table__cell">{{ run.started_display }}</td></tr>
+      <tr class="table__row"><td class="table__cell"><a class="admin__link" href="/admin/runs/{{ run.id }}">{{ run.id }}</a></td><td class="table__cell">{{ run.status_display }}</td><td class="table__cell">{{ run.found_count }}</td><td class="table__cell">{{ run.parsed_count }}</td><td class="table__cell">{{ run.skipped_count }}</td><td class="table__cell">{{ run.new_count }}</td><td class="table__cell">{{ run.error_count }}</td><td class="table__cell">{{ run.dismissed_count }}</td><td class="table__cell">{{ run.started_display }}</td></tr>
      {% endfor %}
    </tbody>
  </table>
--- a/app/version.py
+++ b/app/version.py
@@ -1,3 +1,3 @@
-APP_VERSION = "0.7.5"
-FRONTEND_VERSION = "0.7.5"
-BACKEND_VERSION = "0.7.5"
+APP_VERSION = "0.8.7"
+FRONTEND_VERSION = "0.8.7"
+BACKEND_VERSION = "0.8.7"
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -25,5 +25,15 @@ services:
      postgres:
        condition: service_healthy

+  worker:
+    build: .
+    command: python -m app.worker
+    env_file: .env
+    environment:
+      DATABASE_URL: postgresql+psycopg://${POSTGRES_USER:-miem}:${POSTGRES_PASSWORD:-miem_password}@postgres:5432/${POSTGRES_DB:-miem_workers}
+    depends_on:
+      postgres:
+        condition: service_healthy
+
 volumes:
  postgres_data:
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,7 +1,7 @@
 [project]
 name = "miem-workers"
-version = "0.7.5"
-description = "MIEM employees parser, admin API, and MCP server"
+version = "0.8.7"
+description = "MIEM employees parser, admin API, and web admin"
 requires-python = ">=3.11"
 dependencies = [
  "apscheduler>=3.10.4",
--- a/tests/browser_admin.py
+++ b/tests/browser_admin.py
@@ -0,0 +1,143 @@
+"""Run from the repository root: python tests/browser_admin.py.
+
+Requires Playwright and Chromium. Uses in-memory data and intercepted requests;
+never contacts the production service. Optional BROWSER_EXECUTABLE selects an
+already installed Chromium instead of the Playwright download.
+"""
+import os
+from pathlib import Path
+import sys
+from tempfile import mkdtemp
+from urllib.parse import urlsplit
+
+sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
+
+from fastapi.testclient import TestClient
+from playwright.sync_api import expect, sync_playwright
+from sqlalchemy import create_engine
+from sqlalchemy.orm import sessionmaker
+from sqlalchemy.pool import StaticPool
+
+from app.config import Settings, get_settings
+from app.db import Base, get_db
+from app.main import app
+from app.models import Employee
+from app.security import SESSION_COOKIE, sign_session
+
+
+def main():
+    engine = create_engine("sqlite://", connect_args={"check_same_thread": False}, poolclass=StaticPool)
+    Base.metadata.create_all(engine)
+    factory = sessionmaker(bind=engine)
+    with factory() as db:
+        db.add_all(Employee(profile_key=f"staff:{i}", canonical_url=f"https://www.hse.ru/staff/{i}",
+                            full_name=f"Сотрудник {i:02d}", current_data={"positions": ["Преподаватель"]})
+                   for i in range(30))
+        db.commit()
+
+    def database():
+        with factory() as db:
+            yield db
+
+    settings = Settings(_env_file=None, session_secret="browser-test-secret")
+    app.dependency_overrides[get_db] = database
+    app.dependency_overrides[get_settings] = lambda: settings
+    client = TestClient(app)
+    client.cookies.set(SESSION_COOKIE, sign_session(settings.admin_username, settings))
+    output = Path(mkdtemp(prefix="miem-admin-browser-"))
+    pending = []
+
+    def serve(route):
+        url = urlsplit(route.request.url)
+        if url.path == "/api/crawl-runs/latest":
+            pending.append(route)
+            return
+        response = client.get(url.path + ("?" + url.query if url.query else ""))
+        route.fulfill(status=response.status_code, body=response.content,
+                      content_type=response.headers.get("content-type", "text/plain"))
+
+    try:
+        with sync_playwright() as playwright:
+            browser = playwright.chromium.launch(executable_path=os.getenv("BROWSER_EXECUTABLE"))
+            page = browser.new_page(viewport={"width": 1440, "height": 1000})
+            page.route("**/*", serve)
+            errors = []
+            page.on("pageerror", lambda error: errors.append(str(error)))
+            page.goto("http://miem.test/admin/directory?limit=25&offset=25")
+            expect(page.get_by_role("status")).to_have_text("Показаны 26–30 из 30")
+            assert page.locator("th:visible").count() == 5
+            assert page.locator(".directory__table-wrap").evaluate("e => e.scrollWidth <= e.clientWidth")
+            skip_link = page.get_by_role("link", name="Перейти к содержимому")
+            assert skip_link.evaluate("e => getComputedStyle(e).left === '-999px'")
+            page.keyboard.press("Tab")
+            expect(skip_link).to_be_focused()
+            assert skip_link.evaluate("e => getComputedStyle(e).left === '8px'")
+            page.keyboard.press("Tab")
+            assert skip_link.evaluate("e => getComputedStyle(e).left === '-999px'")
+
+            for width in (1440, 1024, 921, 920, 768, 621, 620, 390, 320):
+                page.set_viewport_size({"width": width, "height": 844})
+                assert page.locator(".directory__filters").evaluate("e => e.scrollWidth <= e.clientWidth")
+                assert page.locator(".directory__input").evaluate_all("""controls => controls.every(control => {
+                  const field = control.closest('.directory__field').getBoundingClientRect();
+                  const rect = control.getBoundingClientRect();
+                  return rect.left >= field.left - 1 && rect.right <= field.right + 1;
+                })""")
+            page.set_viewport_size({"width": 1440, "height": 1000})
+            page.get_by_label("ФИО или ссылка").fill("Сотрудник 00")
+            page.get_by_role("button", name="Применить фильтры").click()
+            expect(page.get_by_role("status")).to_have_text("Показаны 1–1 из 1")
+            assert "offset=" not in page.url
+            trigger = page.get_by_role("button", name="Настроить колонки")
+            trigger.focus()
+            page.keyboard.press("Enter")
+            dialog = page.get_by_role("dialog", name="Отображаемые колонки")
+            expect(dialog).to_be_visible()
+            for _ in range(22):
+                page.keyboard.press("Tab")
+                # Native dialogs may let Tab reach browser chrome, never background controls.
+                assert dialog.evaluate("e => document.activeElement === document.body || e.contains(document.activeElement)")
+            page.keyboard.press("Escape")
+            expect(trigger).to_be_focused()
+            trigger.click()
+            page.get_by_role("button", name="Контакты", exact=True).click()
+            page.keyboard.press("Escape")
+            page.reload()
+            expect(page.locator('th[data-column="email"]')).to_be_visible()
+            page.evaluate("localStorage.clear()")
+            page.goto("http://miem.test/admin/directory")
+            page.screenshot(path=str(output / "desktop.png"), full_page=True)
+            page.set_viewport_size({"width": 390, "height": 844})
+            assert page.evaluate("document.documentElement.scrollWidth <= innerWidth")
+            assert page.locator('th[data-column="full_name"]').evaluate("e => getComputedStyle(e).position") == "sticky"
+            page.screenshot(path=str(output / "mobile.png"), full_page=True)
+            trigger.click()
+            expect(dialog).to_be_visible()
+            assert dialog.evaluate("e => e.getBoundingClientRect().right <= innerWidth")
+            page.screenshot(path=str(output / "mobile-dialog.png"), full_page=True)
+            page.keyboard.press("Escape")
+            page.get_by_label("ФИО или ссылка").fill("Нет совпадений")
+            page.get_by_role("button", name="Применить фильтры").click()
+            expect(page.get_by_text("Сотрудники не найдены", exact=True)).to_be_visible()
+            expect(page.get_by_role("link", name="Сбросить фильтры", exact=True)).to_be_visible()
+            page.goto("http://miem.test/admin")
+            page.wait_for_function("document.querySelector('[data-progress-panel]').getAttribute('aria-busy') === 'true'")
+            pending.pop(0).fulfill(status=503, body="unavailable")
+            expect(page.locator("[data-progress-error]")).to_be_visible()
+            expect(page.get_by_text("Показаны последние полученные данные.", exact=False)).to_be_visible()
+            page.get_by_role("button", name="Повторить").click()
+            page.wait_for_timeout(100)
+            pending.pop(0).fulfill(json={"running": None, "latest": None})
+            expect(page.locator("[data-progress-error]")).to_be_hidden()
+            expect(page.get_by_text("Запусков пока нет. Запустите парсинг.", exact=True)).to_be_visible()
+            assert not errors, errors
+            browser.close()
+    finally:
+        app.dependency_overrides.clear()
+        client.close()
+        engine.dispose()
+    print(f"Browser checks passed; screenshots: {output}")
+
+
+if __name__ == "__main__":
+    main()
--- a/tests/test_academic_degrees.py
+++ b/tests/test_academic_degrees.py
@@ -0,0 +1,26 @@
+from app.services.academic_degrees import academic_degrees
+
+
+def test_academic_degrees_extracts_all_degree_names_without_surrounding_text():
+    data = {
+        "sections": [
+            {
+                "title": "Образование",
+                "year_entries": [
+                    {"text": "2008 — кандидат технических наук, доцент"},
+                    {"text": "2018 — Доктор физико-математических наук"},
+                ],
+            },
+            {
+                "title": "Учёные степени",
+                "table": {"rows": [{"cells": ["Ph.D.", "Доктор физико-математических наук"]}]},
+            },
+            {"title": "Публикации", "items": ["Доктор медицинских наук выступил автором статьи"]},
+        ]
+    }
+
+    assert academic_degrees(data) == [
+        "кандидат технических наук",
+        "Доктор физико-математических наук",
+        "Ph.D.",
+    ]
--- a/tests/test_admin_data.py
+++ b/tests/test_admin_data.py
@@ -268,6 +268,18 @@ def test_list_employees_page_filters_sorts_and_paginates(db_session):
    assert page["limit"] == 50


+def test_catalog_recovers_from_out_of_range_offset(db_session):
+    db_session.add(Employee(profile_key="staff:page", canonical_url="https://www.hse.ru/staff/page", full_name="Page Person"))
+    db_session.commit()
+
+    page = list_employees_page(db_session, q="Page", offset=100, limit=25)
+    assert (page["offset"], page["page"], page["total"]) == (0, 1, 1)
+    assert len(page["employees"]) == 1
+    empty = list_employees_page(db_session, q="missing", offset=100)
+    assert empty["offset"] == 0
+    assert empty["employees"] == []
+
+
 def test_stats_payload_uses_latest_run_new_count(db_session):
    db_session.add(
        Employee(
--- a/tests/test_admin_templates.py
+++ b/tests/test_admin_templates.py
@@ -17,11 +17,11 @@ def test_directory_template_is_russian_and_uses_display_dates():
    template = Path("app/templates/directory.html").read_text(encoding="utf-8")

    assert "Сотрудники" in template
-    assert "Колонки" in template
-    assert "Применить" in template
-    assert "На странице: {{ value }}" in template
+    assert "колонки" in template
+    assert "Применить фильтры" in template
+    assert "Сотрудников на странице" in template
    assert "{% for value in [25, 50, 100] %}" in template
-    assert "Найдено:" in template
+    assert "Показаны" in template
    assert "Новости" in template
    assert "Есть учёная степень" in template
    assert 'data-column="academic_degree"' in template
@@ -29,7 +29,17 @@ def test_directory_template_is_russian_and_uses_display_dates():
    assert "employee.first_seen_display" in template
    assert "employee.last_seen_display" in template
    assert "employee.dismissed_display" in template
-    assert "verification_required" in template
+    assert "verification_required" in template
+    assert '<label class="directory__field">Статус' in template
+    assert 'id="columns-dialog"' in template
+    assert 'aria-controls="columns-dialog"' in template
+    assert 'aria-labelledby="columns-dialog-title"' in template
+    assert "Сбросить фильтры" in template
+    assert 'data-columns-preset="review"' in template
+    assert '<dialog class="columns-modal"' in template
+    assert 'id="columns-dialog"' in template
+    assert 'aria-controls="columns-dialog"' in template
+    assert "Сбросить фильтры" in template
    assert "Directory" not in template
    assert "employees found" not in template

@@ -50,12 +60,9 @@ def test_dashboard_limits_latest_runs_to_five():
 def test_runs_template_links_to_run_detail():
    template = Path("app/templates/runs.html").read_text(encoding="utf-8")

-    assert 'onclick="window.location.href=\'/admin/runs/{{ run.id }}\'"' in template
-    assert "onkeydown=\"if (event.key === 'Enter' || event.key === ' ')" in template
-    assert 'role="link"' in template
-    assert 'tabindex="0"' in template
-    assert 'data-row-href="/admin/runs/{{ run.id }}"' not in template
-    assert '<a class="admin__link" href="/admin/runs/{{ run.id }}">' not in template
+    assert '<a class="admin__link" href="/admin/runs/{{ run.id }}">' in template
+    assert "onclick=" not in template
+    assert 'role="link"' not in template


 def test_run_detail_template_extends_base_and_shows_change_groups():
@@ -86,22 +93,36 @@ def test_dashboard_has_dismissed_status_refresh_action():

    assert 'action="/admin/dismissed/refresh"' in template
    assert "Проверить уволенных" in template
+
+
+def test_progress_template_has_accessible_progressbar():
+    for name in ("dashboard.html", "runs.html"):
+        template = Path(f"app/templates/{name}").read_text(encoding="utf-8")
+        assert 'role="progressbar"' in template
+        assert 'aria-valuemin="0"' in template
+        assert 'aria-valuemax="100"' in template
+        assert "Прогресс парсинга" in template


 def test_dashboard_latest_run_rows_link_to_run_detail():
    template = Path("app/templates/dashboard.html").read_text(encoding="utf-8")

-    assert 'onclick="window.location.href=\'/admin/runs/{{ run.id }}\'"' in template
-    assert "onkeydown=\"if (event.key === 'Enter' || event.key === ' ')" in template
-    assert 'role="link"' in template
-    assert 'tabindex="0"' in template
-    assert 'data-row-href="/admin/runs/{{ run.id }}"' not in template
-    assert '<a class="admin__link" href="/admin/runs/{{ run.id }}">' not in template
+    assert '<a class="admin__link" href="/admin/runs/{{ run.id }}">' in template
+    assert "onclick=" not in template
+    assert 'role="link"' not in template


-def test_admin_js_supports_keyboard_activation_for_clickable_rows():
-    source = Path("app/static/admin.js").read_text(encoding="utf-8")
-
-    assert 'addEventListener("keydown"' in source
-    assert '"Enter"' in source
-    assert '" "' in source
+def test_admin_js_uses_native_dialog_for_column_settings():
+    source = Path("app/static/admin.js").read_text(encoding="utf-8")
+
+    assert "showModal()" in source
+    assert "modal.close()" in source
+    assert 'setAttribute("aria-expanded", "true")' in source
+    assert "data-progress-retry" in source
+
+
+def test_base_template_has_skip_link_and_current_navigation_state():
+    template = Path("app/templates/base.html").read_text(encoding="utf-8")
+
+    assert 'href="#main-content"' in template
+    assert 'aria-current="page"' in template
--- a/tests/test_api.py
+++ b/tests/test_api.py
@@ -0,0 +1,154 @@
+from datetime import datetime, timezone
+from types import SimpleNamespace
+
+from fastapi.testclient import TestClient
+from sqlalchemy import create_engine, select
+from sqlalchemy.exc import OperationalError
+from sqlalchemy.orm import sessionmaker
+from sqlalchemy.pool import StaticPool
+
+from app.config import Settings, get_settings
+from app.db import Base, get_db
+from app.main import app
+from app.models import CrawlRun, CrawlRunEmployeeChange, Employee
+from app.security import SESSION_COOKIE, sign_session
+
+
+def test_health_returns_versions():
+    response = TestClient(app).get("/api/health")
+
+    assert response.status_code == 200
+    assert response.json()["backend_version"] == "0.8.7"
+
+
+def test_api_employees_and_stats_require_admin_session():
+    engine = create_engine("sqlite:///:memory:", connect_args={"check_same_thread": False}, poolclass=StaticPool)
+    Base.metadata.create_all(engine)
+    session_factory = sessionmaker(bind=engine)
+    db = session_factory()
+    employee = Employee(
+        profile_key="staff:alpha",
+        profile_type="staff",
+        profile_id="alpha",
+        canonical_url="https://www.hse.ru/staff/alpha",
+        full_name="Alpha Person",
+        status="active",
+        first_seen_at=datetime.now(timezone.utc),
+        last_seen_at=datetime.now(timezone.utc),
+        current_data={"contacts": {"emails": ["alpha@hse.ru"]}, "sections": []},
+    )
+    run = CrawlRun(source_url="https://miem.hse.ru/persons", status="completed", new_count=1)
+    db.add_all([employee, run])
+    db.commit()
+    db.add(CrawlRunEmployeeChange(
+        crawl_run_id=run.id,
+        employee_id=employee.id,
+        profile_key=employee.profile_key,
+        profile_url=employee.canonical_url,
+        full_name=employee.full_name,
+        change_type="new",
+        profile_available=True,
+        message="added",
+    ))
+    db.commit()
+    run_id = run.id
+    db.close()
+
+    settings = Settings(admin_username="admin", admin_password="password", session_secret="session-secret")
+
+    def override_db():
+        session = session_factory()
+        try:
+            yield session
+        finally:
+            session.close()
+
+    app.dependency_overrides[get_db] = override_db
+    app.dependency_overrides[get_settings] = lambda: settings
+    client = TestClient(app)
+    client.cookies.set(SESSION_COOKIE, sign_session("admin", settings))
+
+    employees = client.get("/api/employees", params={"q": "Alpha", "has_email": True})
+    stats = client.get("/api/stats")
+    run_details = client.get(f"/api/crawl-runs/{run_id}")
+
+    assert employees.status_code == 200
+    assert employees.json()["total"] == 1
+    assert stats.status_code == 200
+    assert stats.json()["new_in_last_run"] == 1
+    assert run_details.status_code == 200
+    assert run_details.json()["changes"]["new"][0]["full_name"] == "Alpha Person"
+    directory = client.get("/admin/directory?q=Alpha&offset=100")
+    assert directory.status_code == 200
+    assert "Показаны 1–1 из 1" in directory.text
+    assert 'name="offset"' not in directory.text
+    assert "Alpha Person" in directory.text
+    empty = client.get("/admin/directory?q=missing&offset=100")
+    assert "Показаны 0–0 из 0" in empty.text
+    assert "Сбросить фильтры" in empty.text
+    app.dependency_overrides.clear()
+
+
+def test_admin_refresh_employee_route_updates_only_requested_employee(monkeypatch):
+    engine = create_engine("sqlite:///:memory:", connect_args={"check_same_thread": False}, poolclass=StaticPool)
+    Base.metadata.create_all(engine)
+    session_factory = sessionmaker(bind=engine)
+    db = session_factory()
+    db.add(Employee(
+        profile_key="org_person:133709486",
+        profile_type="org_person",
+        profile_id="133709486",
+        canonical_url="https://www.hse.ru/org/persons/133709486",
+        full_name="Будков Юрий Алексеевич",
+        status="active",
+    ))
+    db.commit()
+    employee_id = db.scalar(select(Employee.id))
+    db.close()
+
+    settings = Settings(admin_username="admin", admin_password="password", session_secret="session-secret")
+
+    def override_db():
+        session = session_factory()
+        try:
+            yield session
+        finally:
+            session.close()
+
+    calls = []
+
+    def fake_refresh_employee(db, refreshed_employee, route_settings):
+        calls.append((refreshed_employee.id, route_settings))
+        return SimpleNamespace(status="completed")
+
+    app.dependency_overrides[get_db] = override_db
+    app.dependency_overrides[get_settings] = lambda: settings
+    monkeypatch.setattr("app.admin.refresh_employee", fake_refresh_employee)
+    client = TestClient(app)
+    client.cookies.set(SESSION_COOKIE, sign_session("admin", settings))
+
+    response = client.post(f"/admin/employees/{employee_id}/refresh", follow_redirects=False)
+
+    assert response.status_code == 303
+    assert response.headers["location"] == f"/admin/employees/{employee_id}?refresh_status=success"
+    assert calls == [(employee_id, settings)]
+    app.dependency_overrides.clear()
+
+
+def test_directory_database_error_has_retry_and_preserves_filters(db_session, monkeypatch):
+    settings = Settings(_env_file=None, session_secret="test-session-secret")
+    app.dependency_overrides[get_db] = lambda: db_session
+    app.dependency_overrides[get_settings] = lambda: settings
+    def unavailable(*args, **kwargs):
+        raise OperationalError("SELECT", {}, Exception("private database details"))
+    monkeypatch.setattr("app.admin.list_employees_page", unavailable)
+    try:
+        client = TestClient(app)
+        client.cookies.set(SESSION_COOKIE, sign_session(settings.admin_username, settings))
+        response = client.get("/admin/directory?q=Alice")
+        assert response.status_code == 503
+        assert 'href="/admin/directory?q=Alice"' in response.text
+        assert "Повторить загрузку" in response.text
+        assert "private database details" not in response.text
+    finally:
+        app.dependency_overrides.clear()
--- a/tests/test_api_mcp.py
+++ b/tests/test_api_mcp.py
@@ -1,532 +0,0 @@
-import json
-from datetime import datetime, timezone
-from types import SimpleNamespace
-
-from fastapi.testclient import TestClient
-from sqlalchemy import create_engine, select
-from sqlalchemy.orm import sessionmaker
-from sqlalchemy.pool import StaticPool
-
-from app.config import Settings, get_settings
-from app.db import Base, get_db
-from app.main import app
-from app.models import CrawlRun, CrawlRunEmployeeChange, Employee, EmployeePublication
-from app.security import SESSION_COOKIE, sign_session
-
-
-def test_health_returns_versions():
-    client = TestClient(app)
-
-    response = client.get("/api/health")
-
-    assert response.status_code == 200
-    assert response.json()["backend_version"] == "0.7.3"
-
-
-def test_mcp_lists_tools_without_auth_and_ignores_auth_header():
-    engine = create_engine(
-        "sqlite:///:memory:",
-        connect_args={"check_same_thread": False},
-        poolclass=StaticPool,
-    )
-    Base.metadata.create_all(engine)
-    Session = sessionmaker(bind=engine)
-
-    def override_db():
-        session = Session()
-        try:
-            yield session
-        finally:
-            session.close()
-
-    app.dependency_overrides[get_db] = override_db
-    client = TestClient(app)
-
-    without_auth = client.post("/mcp", json={"jsonrpc": "2.0", "id": 1, "method": "tools/list", "params": {}})
-    with_auth = client.post(
-        "/mcp",
-        headers={"Authorization": "Bearer anything"},
-        json={"jsonrpc": "2.0", "id": 1, "method": "tools/list", "params": {}},
-    )
-
-    assert without_auth.status_code == 200
-    assert with_auth.status_code == 200
-    tool_names = {tool["name"] for tool in without_auth.json()["result"]["tools"]}
-    assert "search_employees" in tool_names
-    assert "get_service_info" in tool_names
-    assert "sync_employees" in tool_names
-    assert any(tool["name"] == "get_crawl_run_details" for tool in without_auth.json()["result"]["tools"])
-    assert with_auth.json()["result"]["tools"] == without_auth.json()["result"]["tools"]
-
-    app.dependency_overrides.clear()
-
-
-def test_mcp_search_employees_returns_matching_employee():
-    engine = create_engine(
-        "sqlite:///:memory:",
-        connect_args={"check_same_thread": False},
-        poolclass=StaticPool,
-    )
-    Base.metadata.create_all(engine)
-    Session = sessionmaker(bind=engine)
-    session = Session()
-    session.add(
-        Employee(
-            profile_key="staff:avsergeev",
-            profile_type="staff",
-            profile_id="avsergeev",
-            canonical_url="https://www.hse.ru/staff/avsergeev",
-            full_name="Сергеев Алексей Викторович",
-            status="active",
-            first_seen_at=datetime.now(timezone.utc),
-            last_seen_at=datetime.now(timezone.utc),
-            current_data={"sections": []},
-        )
-    )
-    session.commit()
-    session.close()
-
-    def override_db():
-        db = Session()
-        try:
-            yield db
-        finally:
-            db.close()
-
-    app.dependency_overrides[get_db] = override_db
-    client = TestClient(app)
-
-    response = client.post(
-        "/mcp",
-        json={
-            "jsonrpc": "2.0",
-            "id": 1,
-            "method": "tools/call",
-            "params": {"name": "search_employees", "arguments": {"query": "Сергеев"}},
-        },
-    )
-
-    assert response.status_code == 200
-    assert "Сергеев Алексей Викторович" in response.json()["result"]["content"][0]["text"]
-
-    app.dependency_overrides.clear()
-
-
-def test_mcp_service_info_returns_tools_and_dataset_hash():
-    engine = create_engine(
-        "sqlite:///:memory:",
-        connect_args={"check_same_thread": False},
-        poolclass=StaticPool,
-    )
-    Base.metadata.create_all(engine)
-    Session = sessionmaker(bind=engine)
-    session = Session()
-    session.add(
-        Employee(
-            profile_key="staff:alpha",
-            profile_type="staff",
-            profile_id="alpha",
-            canonical_url="https://www.hse.ru/staff/alpha",
-            full_name="Alpha Person",
-            status="active",
-            current_checksum="a" * 64,
-            current_data={"sections": []},
-        )
-    )
-    session.commit()
-    session.close()
-
-    def override_db():
-        db = Session()
-        try:
-            yield db
-        finally:
-            db.close()
-
-    app.dependency_overrides[get_db] = override_db
-    client = TestClient(app)
-
-    response = client.post(
-        "/mcp",
-        json={"jsonrpc": "2.0", "id": 1, "method": "tools/call", "params": {"name": "get_service_info", "arguments": {}}},
-    )
-
-    assert response.status_code == 200
-    payload = json.loads(response.json()["result"]["content"][0]["text"])
-    assert payload["service_name"] == "miem-employees"
-    assert payload["backend_version"] == "0.7.3"
-    assert payload["dataset"]["hash"]
-    assert any(tool["name"] == "sync_employees" for tool in payload["tools"])
-
-    app.dependency_overrides.clear()
-
-
-def test_mcp_list_employee_publications_prefers_stored_publications_with_fallback():
-    engine = create_engine(
-        "sqlite:///:memory:",
-        connect_args={"check_same_thread": False},
-        poolclass=StaticPool,
-    )
-    Base.metadata.create_all(engine)
-    Session = sessionmaker(bind=engine)
-    session = Session()
-    stored_employee = Employee(
-        profile_key="staff:stored",
-        profile_type="staff",
-        profile_id="stored",
-        canonical_url="https://www.hse.ru/staff/stored",
-        full_name="Stored Person",
-        status="active",
-        current_data={
-            "sections": [
-                {
-                    "type": "publications",
-                    "publications": [{"title": "Old JSON Publication", "url": "https://example.test/old"}],
-                }
-            ]
-        },
-    )
-    fallback_employee = Employee(
-        profile_key="staff:fallback",
-        profile_type="staff",
-        profile_id="fallback",
-        canonical_url="https://www.hse.ru/staff/fallback",
-        full_name="Fallback Person",
-        status="active",
-        current_data={
-            "sections": [
-                {
-                    "type": "publications",
-                    "publications": [{"title": "Fallback Publication", "url": "https://example.test/fallback"}],
-                }
-            ]
-        },
-    )
-    session.add_all([stored_employee, fallback_employee])
-    session.commit()
-    session.add(
-        EmployeePublication(
-            employee_id=stored_employee.id,
-            publication_id="pub-1",
-            title="Stored Publication",
-            year=2024,
-            publication_type="ARTICLE",
-            url="https://publications.hse.ru/view/pub-1",
-            doi_url="https://doi.org/10.1/test",
-            citation_text="Stored Citation",
-            annotation={"ru": "Аннотация", "en": "Abstract"},
-            description={"main": "Stored Citation"},
-            authors=[{"id": "1", "title_ru": "Автор", "is_current_employee": True}],
-            source_hash="a" * 64,
-        )
-    )
-    session.commit()
-    session.close()
-
-    def override_db():
-        db = Session()
-        try:
-            yield db
-        finally:
-            db.close()
-
-    app.dependency_overrides[get_db] = override_db
-    client = TestClient(app)
-
-    stored_response = client.post(
-        "/mcp",
-        json={
-            "jsonrpc": "2.0",
-            "id": 1,
-            "method": "tools/call",
-            "params": {"name": "list_employee_publications", "arguments": {"profile_id_or_url": "stored"}},
-        },
-    )
-    fallback_response = client.post(
-        "/mcp",
-        json={
-            "jsonrpc": "2.0",
-            "id": 2,
-            "method": "tools/call",
-            "params": {"name": "list_employee_publications", "arguments": {"profile_id_or_url": "fallback"}},
-        },
-    )
-
-    stored_payload = json.loads(stored_response.json()["result"]["content"][0]["text"])
-    fallback_payload = json.loads(fallback_response.json()["result"]["content"][0]["text"])
-    assert stored_payload["items"][0]["title"] == "Stored Publication"
-    assert stored_payload["items"][0]["doi_url"] == "https://doi.org/10.1/test"
-    assert stored_payload["items"][0]["annotation"] == {"ru": "Аннотация", "en": "Abstract"}
-    assert stored_payload["items"][0]["authors"] == [{"id": "1", "title_ru": "Автор", "is_current_employee": True}]
-    assert fallback_payload["items"][0]["title"] == "Fallback Publication"
-
-    app.dependency_overrides.clear()
-
-
-def test_mcp_sync_employees_full_empty_and_unknown_hash_modes():
-    engine = create_engine(
-        "sqlite:///:memory:",
-        connect_args={"check_same_thread": False},
-        poolclass=StaticPool,
-    )
-    Base.metadata.create_all(engine)
-    Session = sessionmaker(bind=engine)
-    session = Session()
-    session.add(
-        Employee(
-            profile_key="staff:alpha",
-            profile_type="staff",
-            profile_id="alpha",
-            canonical_url="https://www.hse.ru/staff/alpha",
-            full_name="Alpha Person",
-            status="active",
-            current_checksum="a" * 64,
-            current_data={"sections": [{"type": "paragraphs"}]},
-        )
-    )
-    session.commit()
-    session.close()
-
-    def override_db():
-        db = Session()
-        try:
-            yield db
-        finally:
-            db.close()
-
-    app.dependency_overrides[get_db] = override_db
-    client = TestClient(app)
-
-    full_response = client.post(
-        "/mcp",
-        json={"jsonrpc": "2.0", "id": 1, "method": "tools/call", "params": {"name": "sync_employees", "arguments": {}}},
-    )
-    full_payload = json.loads(full_response.json()["result"]["content"][0]["text"])
-    current_hash = full_payload["to_hash"]
-
-    empty_response = client.post(
-        "/mcp",
-        json={
-            "jsonrpc": "2.0",
-            "id": 2,
-            "method": "tools/call",
-            "params": {"name": "sync_employees", "arguments": {"client_hash": current_hash}},
-        },
-    )
-    empty_payload = json.loads(empty_response.json()["result"]["content"][0]["text"])
-
-    unknown_response = client.post(
-        "/mcp",
-        json={
-            "jsonrpc": "2.0",
-            "id": 3,
-            "method": "tools/call",
-            "params": {"name": "sync_employees", "arguments": {"client_hash": "missing"}},
-        },
-    )
-    unknown_payload = json.loads(unknown_response.json()["result"]["content"][0]["text"])
-
-    assert full_payload["mode"] == "full"
-    assert full_payload["items"][0]["data"] == {"sections": [{"type": "paragraphs"}]}
-    assert empty_payload["mode"] == "delta"
-    assert empty_payload["changes"] == {"added": [], "updated": [], "dismissed": [], "removed": []}
-    assert unknown_payload["mode"] == "full"
-    assert unknown_payload["reason"] == "unknown_client_hash"
-
-    app.dependency_overrides.clear()
-
-
-def test_mcp_get_crawl_run_details_returns_changes():
-    engine = create_engine(
-        "sqlite:///:memory:",
-        connect_args={"check_same_thread": False},
-        poolclass=StaticPool,
-    )
-    Base.metadata.create_all(engine)
-    Session = sessionmaker(bind=engine)
-    session = Session()
-    run = CrawlRun(source_url="https://miem.hse.ru/persons", status="completed", new_count=1)
-    employee = Employee(
-        profile_key="staff:new",
-        profile_type="staff",
-        profile_id="new",
-        canonical_url="https://www.hse.ru/staff/new",
-        full_name="New Person",
-        status="active",
-        first_seen_at=datetime.now(timezone.utc),
-        last_seen_at=datetime.now(timezone.utc),
-    )
-    session.add_all([run, employee])
-    session.commit()
-    session.add(
-        CrawlRunEmployeeChange(
-            crawl_run_id=run.id,
-            employee_id=employee.id,
-            profile_key=employee.profile_key,
-            profile_url=employee.canonical_url,
-            full_name=employee.full_name,
-            change_type="new",
-            profile_available=True,
-            message="added",
-        )
-    )
-    session.commit()
-    run_id = run.id
-    session.close()
-
-    def override_db():
-        db = Session()
-        try:
-            yield db
-        finally:
-            db.close()
-
-    app.dependency_overrides[get_db] = override_db
-    client = TestClient(app)
-
-    response = client.post(
-        "/mcp",
-        json={
-            "jsonrpc": "2.0",
-            "id": 1,
-            "method": "tools/call",
-            "params": {"name": "get_crawl_run_details", "arguments": {"run_id": run_id}},
-        },
-    )
-
-    assert response.status_code == 200
-    text = response.json()["result"]["content"][0]["text"]
-    assert "New Person" in text
-    assert "changes_detail_available" in text
-
-    app.dependency_overrides.clear()
-
-
-def test_mcp_protected_resource_metadata_route_is_removed():
-    client = TestClient(app)
-
-    response = client.get("/.well-known/oauth-protected-resource")
-
-    assert response.status_code == 404
-
-
-def test_api_employees_and_stats_require_admin_session():
-    engine = create_engine(
-        "sqlite:///:memory:",
-        connect_args={"check_same_thread": False},
-        poolclass=StaticPool,
-    )
-    Base.metadata.create_all(engine)
-    Session = sessionmaker(bind=engine)
-    db = Session()
-    db.add(
-        Employee(
-            profile_key="staff:alpha",
-            profile_type="staff",
-            profile_id="alpha",
-            canonical_url="https://www.hse.ru/staff/alpha",
-            full_name="Alpha Person",
-            status="active",
-            first_seen_at=datetime.now(timezone.utc),
-            last_seen_at=datetime.now(timezone.utc),
-            current_data={"contacts": {"emails": ["alpha@hse.ru"]}, "sections": []},
-        )
-    )
-    run = CrawlRun(source_url="https://miem.hse.ru/persons", status="completed", new_count=1)
-    db.add(run)
-    db.commit()
-    db.add(
-        CrawlRunEmployeeChange(
-            crawl_run_id=run.id,
-            employee_id=1,
-            profile_key="staff:alpha",
-            profile_url="https://www.hse.ru/staff/alpha",
-            full_name="Alpha Person",
-            change_type="new",
-            profile_available=True,
-            message="added",
-        )
-    )
-    db.commit()
-    run_id = run.id
-    db.close()
-
-    settings = Settings(admin_username="admin", admin_password="password", session_secret="session-secret")
-
-    def override_db():
-        session = Session()
-        try:
-            yield session
-        finally:
-            session.close()
-
-    app.dependency_overrides[get_db] = override_db
-    app.dependency_overrides[get_settings] = lambda: settings
-    client = TestClient(app)
-    client.cookies.set(SESSION_COOKIE, sign_session("admin", settings))
-
-    employees = client.get("/api/employees", params={"q": "Alpha", "has_email": True})
-    stats = client.get("/api/stats")
-    run_details = client.get(f"/api/crawl-runs/{run_id}")
-
-    assert employees.status_code == 200
-    assert employees.json()["total"] == 1
-    assert stats.status_code == 200
-    assert stats.json()["new_in_last_run"] == 1
-    assert run_details.status_code == 200
-    assert run_details.json()["changes"]["new"][0]["full_name"] == "Alpha Person"
-
-    app.dependency_overrides.clear()
-
-
-def test_admin_refresh_employee_route_updates_only_requested_employee(monkeypatch):
-    engine = create_engine(
-        "sqlite:///:memory:",
-        connect_args={"check_same_thread": False},
-        poolclass=StaticPool,
-    )
-    Base.metadata.create_all(engine)
-    Session = sessionmaker(bind=engine)
-    db = Session()
-    db.add(
-        Employee(
-            profile_key="org_person:133709486",
-            profile_type="org_person",
-            profile_id="133709486",
-            canonical_url="https://www.hse.ru/org/persons/133709486",
-            full_name="Будков Юрий Алексеевич",
-            status="active",
-        )
-    )
-    db.commit()
-    employee_id = db.scalar(select(Employee.id))
-    db.close()
-
-    settings = Settings(admin_username="admin", admin_password="password", session_secret="session-secret")
-
-    def override_db():
-        session = Session()
-        try:
-            yield session
-        finally:
-            session.close()
-
-    calls = []
-
-    def fake_refresh_employee(db, refreshed_employee, route_settings):
-        calls.append((refreshed_employee.id, route_settings))
-        return SimpleNamespace(status="completed")
-
-    app.dependency_overrides[get_db] = override_db
-    app.dependency_overrides[get_settings] = lambda: settings
-    monkeypatch.setattr("app.admin.refresh_employee", fake_refresh_employee)
-    client = TestClient(app)
-    client.cookies.set(SESSION_COOKIE, sign_session("admin", settings))
-
-    response = client.post(f"/admin/employees/{employee_id}/refresh", follow_redirects=False)
-
-    assert response.status_code == 303
-    assert response.headers["location"] == f"/admin/employees/{employee_id}?refresh_status=success"
-    assert calls == [(employee_id, settings)]
-
-    app.dependency_overrides.clear()
--- a/tests/test_config.py
+++ b/tests/test_config.py
@@ -11,3 +11,12 @@ def test_numeric_crawl_limit_is_parsed():
    settings = Settings(crawl_limit="25")

    assert settings.crawl_limit == 25
+
+
+def test_dismissal_safety_settings_have_conservative_defaults(monkeypatch):
+    monkeypatch.delenv("DISMISSAL_CONFIRMATION_RUNS", raising=False)
+    monkeypatch.delenv("MAX_AUTO_DISMISSALS_PER_RUN", raising=False)
+    settings = Settings(_env_file=None)
+
+    assert settings.dismissal_confirmation_runs == 3
+    assert settings.max_auto_dismissals_per_run == 25
--- a/tests/test_dataset_versions.py
+++ b/tests/test_dataset_versions.py
@@ -1,88 +0,0 @@
-from datetime import datetime, timezone
-
-from app.models import Employee
-from app.services.dataset_versions import get_or_create_current_version, sync_employees_payload
-
-
-def _employee(profile_key: str, checksum: str, *, status: str = "active") -> Employee:
-    return Employee(
-        profile_key=profile_key,
-        profile_type=profile_key.split(":", 1)[0],
-        profile_id=profile_key.split(":", 1)[1],
-        canonical_url=f"https://www.hse.ru/{profile_key}",
-        full_name=profile_key,
-        status=status,
-        first_seen_at=datetime.now(timezone.utc),
-        last_seen_at=datetime.now(timezone.utc),
-        current_data={"profile_key": profile_key},
-        current_checksum=checksum,
-    )
-
-
-def test_dataset_version_hash_is_stable_for_same_employee_state(db_session):
-    db_session.add(_employee("staff:alpha", "a" * 64))
-    db_session.commit()
-
-    first = get_or_create_current_version(db_session)
-    db_session.commit()
-    second = get_or_create_current_version(db_session)
-
-    assert second.id == first.id
-    assert second.hash == first.hash
-    assert second.employee_count == 1
-
-
-def test_dataset_version_hash_changes_when_employee_checksum_changes(db_session):
-    employee = _employee("staff:alpha", "a" * 64)
-    db_session.add(employee)
-    db_session.commit()
-    first = get_or_create_current_version(db_session)
-    db_session.commit()
-
-    employee.current_checksum = "b" * 64
-    db_session.commit()
-    second = get_or_create_current_version(db_session)
-
-    assert second.hash != first.hash
-    assert second.previous_hash == first.hash
-
-
-def test_sync_employees_diff_spans_multiple_intermediate_versions(db_session):
-    alpha = _employee("staff:alpha", "a" * 64)
-    db_session.add(alpha)
-    db_session.commit()
-    first = get_or_create_current_version(db_session)
-    db_session.commit()
-
-    beta = _employee("staff:beta", "b" * 64)
-    db_session.add(beta)
-    db_session.commit()
-    get_or_create_current_version(db_session)
-    db_session.commit()
-
-    alpha.current_checksum = "c" * 64
-    alpha.current_data = {"profile_key": "staff:alpha", "changed": True}
-    db_session.commit()
-
-    payload = sync_employees_payload(db_session, client_hash=first.hash, include_data=False)
-
-    assert payload["mode"] == "delta"
-    assert [item["profile_key"] for item in payload["changes"]["added"]] == ["staff:beta"]
-    assert [item["profile_key"] for item in payload["changes"]["updated"]] == ["staff:alpha"]
-    assert payload["changes"]["dismissed"] == []
-    assert payload["changes"]["removed"] == []
-
-
-def test_sync_employees_reports_dismissed_as_tombstone(db_session):
-    alpha = _employee("staff:alpha", "a" * 64)
-    db_session.add(alpha)
-    db_session.commit()
-    first = get_or_create_current_version(db_session)
-    db_session.commit()
-
-    alpha.status = "dismissed"
-    db_session.commit()
-    payload = sync_employees_payload(db_session, client_hash=first.hash, include_data=False)
-
-    assert payload["changes"]["dismissed"][0]["profile_key"] == "staff:alpha"
-    assert payload["changes"]["dismissed"][0]["status"] == "dismissed"
Author	SHA1	Message	Date
agent	09fa15af0b	Fix release upload over SSH	2026-09-14 18:13:51 +03:00
admin	f6a01500d7	Merge pull request 'fix: upload deployment archive to user home' (#44 ) from fix/cd-release-upload-path into main Some checks failed CI / test (push) Successful in 9m19s Details CI / deploy (push) Failing after 15s Details Reviewed-on: #44	2026-09-14 14:33:14 +00:00
agent	ec7aec310a	fix: stage deploy archive in user home All checks were successful CI / test (pull_request) Successful in 9m22s Details CI / deploy (pull_request) Has been skipped Details	2026-09-14 17:23:29 +03:00
admin	0f91ed0d8f	Merge pull request 'fix: repair admin directory filter layout' (#43 ) from fix/admin-directory-filters into main Some checks failed CI / test (push) Successful in 9m23s Details CI / deploy (push) Failing after 14s Details Reviewed-on: #43	2026-09-14 13:58:41 +00:00
agent	1e9cc50a73	fix: keep filter select values readable All checks were successful CI / test (pull_request) Successful in 9m20s Details CI / deploy (pull_request) Has been skipped Details	2026-09-14 15:28:02 +03:00
agent	10d2cf0101	fix: repair admin directory filter layout All checks were successful CI / test (pull_request) Successful in 9m21s Details CI / deploy (pull_request) Has been skipped Details	2026-09-14 15:24:41 +03:00
admin	44949996bc	Merge pull request 'fix: configure dismissal safeguards and remove obsolete endpoint references' (#42 ) from refactor/remove-mcp-mentions into main Some checks failed CI / test (push) Successful in 9m20s Details CI / deploy (push) Failing after 42s Details Reviewed-on: #42	2026-09-14 10:26:12 +00:00
admin	baad49b976	docs: record dismissal config fix All checks were successful CI / test (pull_request) Successful in 9m19s Details CI / deploy (pull_request) Has been skipped Details	2026-09-14 12:53:01 +03:00
admin	ceca2017cb	fix: wire dismissal safety settings All checks were successful CI / test (pull_request) Successful in 9m20s Details CI / deploy (pull_request) Has been skipped Details	2026-09-14 12:48:47 +03:00
admin	83e726efe5	docs: align service version in README All checks were successful CI / test (pull_request) Successful in 9m18s Details CI / deploy (pull_request) Has been skipped Details	2026-09-14 12:41:04 +03:00
admin	7d6d8759a5	refactor: remove obsolete endpoint references All checks were successful CI / test (pull_request) Successful in 9m23s Details CI / deploy (pull_request) Has been skipped Details	2026-09-14 12:39:06 +03:00
admin	5d89ae3891	Merge pull request 'ci: add protected production deploy workflow' (#40 ) from chore/production-cd into main All checks were successful CI / test (push) Successful in 12m58s Details CI / deploy (push) Successful in 1m18s Details Reviewed-on: #40	2026-09-11 12:17:48 +00:00
admin	880ac6c53e	ci: gate production deploy on tests Some checks failed CI / test (pull_request) Failing after 1s Details CI / deploy (pull_request) Has been skipped Details	2026-09-11 13:26:59 +03:00
admin	2d19eef9c2	ci: add protected production deploy workflow Some checks failed CI / test (pull_request) Failing after 1s Details	2026-09-11 13:16:43 +03:00
admin	a1d60984e3	Merge pull request 'ci: run tests on main and pull requests' (#39 ) from chore/ci-cd-production into main Some checks failed CI / test (push) Has been cancelled Details Reviewed-on: #39	2026-09-11 10:06:21 +00:00
admin	d5f41e015f	ci: run tests on pushes and pull requests Some checks failed CI / test (pull_request) Failing after 12s Details	2026-09-11 13:04:54 +03:00
admin	e1ba77c337	Merge pull request 'fix: complete admin catalog UX and recovery' (#38 ) from feature/admin-ux-hardening into main Reviewed-on: #38	2026-09-11 09:50:03 +00:00
admin	a989c5c53c	docs: align version and test instructions	2026-09-11 12:35:02 +03:00
admin	a82930027d	fix: complete catalog pagination and recovery states	2026-09-11 11:25:59 +03:00
admin	78493b9248	fix: complete admin UX review findings	2026-08-28 16:01:00 +03:00
admin	5d3fb68c2f	fix: harden admin UX and accessibility	2026-08-28 12:11:52 +03:00
admin	dc060985a3	Merge pull request 'fix: display all academic degrees' (#36 ) from fix/academic-degree-display into main Reviewed-on: #36	2026-08-27 14:34:17 +00:00
admin	1d8a854a12	chore: align package version	2026-08-27 17:32:04 +03:00
admin	d07711f665	fix: display all academic degrees	2026-08-27 17:29:16 +03:00
admin	cf810f9cad	Merge pull request 'refactor: remove unused data integration' (#35 ) from fix/remove-mcp into main Reviewed-on: #35	2026-08-27 14:12:37 +00:00
admin	e142138c1b	refactor: remove dataset sync leftovers	2026-08-27 17:08:33 +03:00
admin	c031fb9f4a	refactor: remove unused data integration	2026-08-27 17:04:03 +03:00
admin	70a4719bab	Merge pull request 'fix: restore scheduled crawl worker' (#34 ) from fix/restore-scheduled-worker into main Reviewed-on: #34	2026-08-26 15:23:46 +00:00
admin	3bdb8b1d90	docs: add scheduled worker changelog	2026-08-26 18:22:45 +03:00
admin	f3fb714126	fix: restore scheduled crawl worker	2026-08-26 18:21:04 +03:00
admin	3ed1dd8832	Merge pull request 'fix: run only API in compose' (#33 ) from fix/api-only-compose into main Reviewed-on: #33	2026-08-26 15:16:33 +00:00