Add automation for lts testing: clickhouse-odbc, grafana, superset, dbeaver

[Selfeer]

Add LTS testing skills and automation framework for web/desktop UI tools

Summary

• Introduce a new lts/ folder structure for LTS (Long-Term Support) testing automation covering four integration targets: lts/clickhouse-odbc, lts/grafana, lts/superset, and lts/dbeaver
• Each sub-suite follows the established TestFlows test structure (modeled after altinity/clickhouse-regression and altinity/clickhouse-grafana) with a top-level regression.py, requirements/, steps/, and tests/ layout
• Add AGENT.md documenting proper test structure conventions, including: requirements definition, requirement-to-test-scenario mapping (ideally 1:1, or one higher-level requirement per feature), test steps organization, and test scenario authoring

Scope per target

Target	Type	Automation approach
clickhouse-odbc	Driver/API	Fully automated via TestFlows
grafana	Web UI	Fully automated (Selenium/Playwright), following existing altinity/clickhouse-grafana UI test patterns
superset	Web UI	Fully automated (Selenium/Playwright) web UI tests
dbeaver	Desktop UI	Semi-automated: agent-driven or manual steps where desktop GUI automation is limited

Test structure (per AGENT.md)

lts/<target>/
├── regression.py          # Main entry point, xfails/ffails, argparser
├── requirements/
│   ├── requirements.md    # Human-readable SRS requirements
│   └── requirements.py    # TestFlows requirement objects
├── steps/                 # Reusable test step functions (actions, UI interactions)
│   └── ...
└── tests/                 # Test scenarios (map 1:1 to requirements where possible)
    └── ...

Key conventions

• Requirements: Each requirement has a unique ID and is traceable to one or more test scenarios
• Requirement-to-scenario mapping: Ideally 1:1 for granular requirements; one higher-level requirement maps to a feature-level test suite
• Test steps: Reusable, composable step functions (Given/When/Then) shared across scenarios
• Test scenarios: Self-contained, independently runnable tests that explicitly link back to their requirement(s)

Test plan

• Verify lts/ folder structure is created with all four sub-suites
• Confirm each sub-suite has regression.py, requirements/, steps/, tests/
• Validate AGENT.md documents the test structure, requirements mapping, and conventions
• Ensure lts/grafana follows the pattern from altinity/clickhouse-grafana UI tests
• Review requirement-to-scenario traceability in each suite