---

name: ha-integration-maintainer description: Home Assistant + HACS maintenance for the Heat Pump Predictor integration. Use when adding or updating config flow, coordinator, entities, services, translations, recorder/statistics compatibility, dashboards, or HACS release assets. Trigger for any HA/HACS changes to custom_components/heat_pump_predictor/.

Home Assistant Integration Maintainer (Heat Pump Predictor)

Use this skill whenever modifying or validating the Heat Pump Predictor integration for Home Assistant (HACS-compatible). Keep outputs concise; prefer checklists and ready snippets.

Scope and Triggers

• Changes to config flow, coordinator, entities, services, translations, recorder/statistics, dashboards, or HACS packaging.
• Debugging UpdateFailed, missing entities, translation or manifest issues, recorder/state_class warnings, or HACS validation.
• Preparing releases (version bumps, docs alignment) for heat_pump_predictor.

Quick Checklist (per change)

• Config flow: async_step_user with voluptuous; validate 3 inputs (energy total_increasing, running binary, temperature sensor); surface errors via errors dict.
• Coordinator: DataUpdateCoordinator with 5m interval; call async_config_entry_first_refresh; raise UpdateFailed on API/state errors.
• Entities: subclass CoordinatorEntity + platform class; set _attr_has_entity_name = True; main feature entities set _attr_name = None; use translation_key; unique_id per entry + key; device_info from coordinator.
• Translations: add/align strings.json keys (flow, options, entities); no hard-coded English in code.
• Recorder/Statistics: correct device_class, unit, state_class (energy = total_increasing, power = measurement, duty = measurement, precision per metric); avoid unavailable/unknown math.
• Services: register calculate_energy with schema; validate entry id + temperature; async-only I/O.
• Manifest/HACS: manifest fields (domain, name, version SemVer, codeowners, documentation, issue_tracker, config_flow true, integration_type, iot_class); hacs.json present; README aligned; bump version on release.
• Logging/diagnostics: lazy formatting; log once per failure path; do not swallow exceptions.
• Tests: use MockConfigEntry + hass.config_entries.async_setup; assert hass.states.get; avoid direct module import; snapshot complex outputs when needed.

Core Patterns and Snippets

• Config Flow (user step)

  • async_step_user: schema requires energy_sensor, running_sensor, temperature_sensor; validate entity domains/types; return self.async_create_entry on success; set errors["base"] for CannotConnect/InvalidAuth/unknown.
  • Use voluptuous selectors if extending; keep async for entity lookups.

• Coordinator Setup

  • Create DataUpdateCoordinator(hass, LOGGER, name=DOMAIN, update_interval=timedelta(minutes=5), update_method=_async_update_data).
  • Call await coordinator.async_config_entry_first_refresh() before forwarding platforms.
  • In _async_update_data, raise UpdateFailed on network/parse/state issues; never block.

• Entity Template

  • class MyEntity(CoordinatorEntity, SensorEntity): _attr_has_entity_name = True; attr_name = None for main feature entities; translation_key for additional ones; unique_id f"{entry.entry_id}{description.key}".
  • Use entity_description with value_fn; in available, rely on coordinator.last_update_success.

• Translations

  • Use translation_key matching strings.json; avoid hard-coded English names; strings.json holds entity/device/flow text.

• Recorder/Statistics

  • Energy: device_class=energy, unit=kWh, state_class=total_increasing.
  • Power: device_class=power, unit=W, state_class=measurement.
  • Duty cycle: unit=%, state_class=measurement; round sensibly.
  • Guard against None/unavailable before math; return None when data missing.

• Services

  • calculate_energy: validate temperature bucket and config_entry_id; async handler using coordinator/data manager; catch and log predictable errors, re-raise HomeAssistantError for user-facing issues.

• Dashboards (ApexCharts helpers)

  • Curve sensors expose attributes.data with entries like {temp, power_overall, power_running} or {temp, duty_cycle} or {temp, energy}; data_generator maps to [temp, value].
  • Recommend enabling only relevant bucket sensors (disabled by default) to reduce entity load.

Common Pitfalls

• Missing _attr_has_entity_name = True or hard-coded English names; fix via translation_key and _attr_name = None for main entities.
• Forgetting async_config_entry_first_refresh leading to None data on startup.
• Swallowing exceptions in update_method; always raise UpdateFailed.
• Incorrect state_class/unit/device_class causing recorder warnings.
• Unique_id collisions; ensure entry-scoped unique_id.
• Not bumping manifest version for releases; HACS will complain.

Release and Validation Steps

• Update manifest version (SemVer) and ensure documentation/issue_tracker URLs are present.
• Confirm hacs.json name set; README matches behavior.
• Run tests (pytest) if available; lint if configured.
• Validate strings.json is valid JSON and covers translation_keys used.
• Tag/release per repo workflow; ensure assets in custom_components/heat_pump_predictor/ only.

References (read as needed)

• Project README for sensor expectations and ApexCharts examples.
• custom_components/heat_pump_predictor/const.py for domain, keys, and intervals.
• copilot/python-ha-integration.md for coding conventions (imports, async, logging, entities, constants).