- add TraceInterceptor that reads x-trace-id header or generates a UUID
- add AsyncLocalStorage store in common/trace-context for request-scoped trace ID
- replace NestJS Logger with TraceLogger (extends ConsoleLogger) that injects trace ID into log context
- register TraceInterceptor globally before LoggingInterceptor
- move HealthController into HealthModule so global interceptors apply to /healthz
- omit body/response from log lines when empty or null
- KeyDescriptor now accepts optional login field alongside keyFile
- when login is present, clicks login/password auth method, fills email+password fields, and submits
- file-key flow unchanged; validation ensures exactly one strategy is present
- system Chromium path from PLAYWRIGHT_CHROMIUM_EXECUTABLE_PATH was not forwarded
to the launch options, causing 'executable doesn't exist' errors in Docker
- add PaginationQueryDto<TOrderBy> generic with orderBy and orderDir fields
- add SessionOrderBy, EnvironmentOrderBy, ScenarioOrderBy type aliases
- update session, environment, scenario services and controllers to use typed pagination
- update MCP list_sessions, list_environments, list_scenarios tools to expose orderBy/orderDir
- update tests for all list endpoints to cover page, limit, ordering, and invalid param rejection
- add HttpExceptionFilter logging BadRequestException at warn and InternalServerErrorException at error with cause chain
- add LoggingInterceptor logging request/response pairs at debug level with method, path, body, status, and duration
- extract CodeExecutorService into standalone CodeExecutorModule imported by BrowserModule and McpModule
- thread { cause: err } into all catch blocks across auth, browser, and code-executor services
- automates file-key login on the ID portal using Playwright/Chromium
- captures token, cookies, and localStorage after successful redirect
- stores session data in SQLite via TypeORM (better-sqlite3)
- sessions are keyed by sessionName (auto-generated UUID if not provided)
- login URL, cabinet redirect URL, keys dir, and DB path are all configurable via env