This folder contains local developer helpers and repository-maintenance scripts. They are not imported by the add-on at runtime.
assess.ps1runs focused assessment checks against a live environment.api_surface_smoke.pyruns the HTTP API in-process with FastAPITestClient, seeded SQLite state, and local stubs so the API surface can be regression-tested without deploying to Home Assistant. It covers health, dev status, topology/history, partitions, routing, stale links, network data, assessment, chat telemetry, and the prompt corpus through/v1/chat/turn.
Run it locally with PYTHONPATH=addons/thread-observability/app/src python scripts/api_surface_smoke.py.
chat-smoke.ps1exercises chat flows against the add-on, including persisted transcript inspection through/v1/chat/transcript/{conversation_id}when transcript persistence is enabled. Semantic pass/fail now comes fromchat_smoke_evaluator.py, which calls an OpenAI-compatible reviewer model usingTHREAD_OBS_SMOKE_EVAL_PROVIDER,THREAD_OBS_SMOKE_EVAL_BASE_URL,THREAD_OBS_SMOKE_EVAL_MODEL,THREAD_OBS_SMOKE_EVAL_API_KEY, and optionalTHREAD_OBS_SMOKE_EVAL_TEMPERATURE.direct_chat_prompt_smoke.pyis the prompt-focused Python smoke path: it takes a list of pre-authored likely user queries, runs them through the realdirect_chat_turnorchestration in-process, uses AI to judge each final answer, and when a case fails asks AI to review the full transcript including the embedded reviewer/audit step. It also supports model comparison with--targets <json>.dashboard-loop.ps1repeats dashboard-oriented checks during live validation.generate_mcp_reference.pyregenerates the MCP reference documentation from the live tool registry.test_real_logs.pyis an ad hoc OTBR parser smoke helper for quickly checking a few real log lines outside the automated test suite.