agentsclimarketplace

Ha automation solution

Skill nolte/claude-home-assistant/skills/ha-automation-solution

Plans and orchestrates a complete Home Assistant YAML solution from a result-oriented requirement, so the user never has to pick which authoring skill to use. Decomposes the requirement into the minimal combination of artifacts across the ha-automation skill family, presents a dependency-ordered artifact plan for approval, then dispatches ha-automation-author, ha-helper-scaffold, ha-derived-sensor-author, and ha-blueprint-scaffold in order — threading entity_ids between steps — and flags requirements that actually need a custom integration. Activate on "I want my heat pump's daily energy on the dashboard and an alert when it's high", "set up presence-based lighting that only runs in the evening", "baue mir eine Lösung, die…", "richte… ein". Do not activate for a single clear artifact (let the owning skill handle it), Python custom integrations (ha-integration-scaffold), or deploying to a live HA instance. Supports resume on re-invocation.From its SKILL.md

Install
npx -y skills add nolte/claude-home-assistant --skill ha-automation-solution

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

9.8 KB, as published. Nobody here has run it

HA Automation Solution

Spec: spec/claude/ha-automation-solution/en.md (EN canonical) / spec/claude/ha-automation-solution/de.md (DE translation).

This skill is the front door to the ha-automation skill family. It does not generate any artifact itself — it decomposes the requirement, plans the combination, and dispatches the owning authoring skills, each of which owns its generation and spec conformance.

Why this is a skill, not an agent

  • Plan-before-generate gate — the artifact plan must be presented and explicitly approved before any generation; that human-visible gate is core to the contract and an agent's fire-and-forget shape would lose it.
  • Mid-flow interactivity — clarifying questions, plan confirmation, and the "this actually needs a custom integration" decision are per-run dialogues.
  • Orchestrator that dispatches other skills — the skill-orchestrates-skill default (see skill-vs-agent) keeps the entry point in skill form, like ha-blueprint-scaffold dispatching ha-blueprint-author.
  • Counter-dimension considered: the per-artifact generation could run as parallel agents, but the plan approval and the entity_id threading must stay visible in the user's context; skill wins.

When this skill activates

Use this skill when the user describes a result that likely needs more than one artifact, and should not have to know which authoring skill produces what.

When NOT to activate

  • a single clear artifact (one automation, one helper, one sensor, one blueprint) → let the owning skill activate directly
  • a Python custom integration (own device/cloud protocol, polling, config flow) → ha-integration-scaffold
  • deploying/importing into a running HA instance → out of scope (generation only)

Hard rules

  1. Never generate inline. Every artifact is produced by its owning skill, resolved at runtime from the live ha-automation family inventory (see Runtime skill resolution) rather than a frozen name list — the skill names in the decomposition heuristic are illustrative anchors. This skill plans and dispatches; it does not write artifacts.
  2. Plan before generate. Always present the dependency-ordered artifact plan and wait for explicit approval before dispatching anything.
  3. One requirement, one run. No multi-requirement batches.
  4. Minimal artifacts. Decompose to the fewest artifacts that satisfy the requirement; never add a helper or sensor a single artifact already covers.
  5. Thread identities. Dispatch in dependency order and pass the entity_ids/identifiers produced in earlier steps as inputs to dependent steps. Keep all names consistent per spec/ha/naming-conventions/en.md.
  6. Stop on NEEDS-WORK. If a dispatched skill returns NEEDS-WORK, stop and report — do not build a dependent artifact on an unfinished predecessor.
  7. Recognize integration-shaped work. When the requirement needs a custom integration rather than YAML, say so in the plan and point at ha-integration-scaffold instead of forcing it into helpers/templates.
  8. Verify HA internals against the official docs (see spec/ha/upstream-docs-verification/en.md).

Inputs

FieldRequiredDefaultNotes
requirementyes—The desired result, in prose
target_dir / target_filenoworking dirpassed through to dispatched skills
known_sourcesnoasked when neededexisting entity_ids to use as sources

Runtime skill resolution

Resolve the owning skill for each artifact at runtime, by matching the requirement against the live inventory of this plugin's ha-automation family (ha-automation-author, ha-helper-scaffold, ha-derived-sensor-author, ha-blueprint-scaffold) — read each candidate's stated responsibility from your available-skills registry, or, when running inside the plugin source tree, Glob skills/ha-*/SKILL.md and read its description:. Match on responsibility, not on a remembered name. The decomposition heuristic below is an illustrative anchor of the typical mappings, not an authoritative or exhaustive list: re-resolve against the current inventory on every run, so a skill newly added to (or renamed within) the family is dispatchable immediately and a removed one is not — without editing this skill (the runtime-lookup pattern of issue-orchestrate). If you genuinely cannot enumerate the live inventory, fall back to the anchor table and note the degraded resolution.

Decomposition heuristic (requirement → artifact type → skill)

The requirement needs…ArtifactOwning skill
a measured/derived value (rate, smoothing, integral, aggregate, threshold, trend, consumption cycle, probability)derived/statistical sensorha-derived-sensor-author
a manually/automation-held state, mode switch, countdown, weekly planinput_* / counter / timer / scheduleha-helper-scaffold
event→action logicautomationha-automation-author
a reusable manually-callable action sequencescriptha-automation-author
a one-shot snapshot of entity statessceneha-automation-author
a free-form Jinja-computed entitygeneric templateha-automation-author
an HTTP / shell / python escape-hatchrest_command / shell_command / python_scriptha-automation-author
a reusable, parameterized, shareable patternblueprintha-blueprint-scaffold
an own device/cloud protocol, polling, config flowcustom integrationout of scope → ha-integration-scaffold

Workflow

1) Clarify

First gauge requirement confidence. When the requirement is clearly specified, use the lightweight path: ask 1–3 targeted questions (which source entity, which threshold, which time windows) before planning. When it is below a confidence threshold (vague trigger, unnamed entities, unclear scope), dispatch requirements-elicit (from the nolte-shared plugin; when it is not installed, reach the same rigor via a structured series of targeted questions) first and plan against the confirmed requirement artifact — mirroring the issue-orchestrate upstream gate — instead of decomposing a fuzzy requirement against weak understanding. Do not plan on guesses.

2) Plan

Decompose into a dependency-ordered artifact plan and present it as a table:

| # | Artifact / entity_id | Type | Skill | Depends on | Purpose |
|---|---|---|---|---|---|
| 1 | sensor.heat_pump_energy | integration | ha-derived-sensor-author | — | Riemann power→energy |
| 2 | sensor.heat_pump_energy_daily | utility_meter | ha-derived-sensor-author | #1 | daily cycle |
| 3 | binary_sensor.heat_pump_energy_high | threshold | ha-derived-sensor-author | #2 | > 10 kWh |
| 4 | automation.heat_pump_energy_alert | automation | ha-automation-author | #3 | notify |

State any custom-integration finding here. Wait for explicit approval.

3) Dispatch

Invoke each owning skill in plan order, passing the entity_ids resolved in earlier steps as inputs to the dependent steps. After each, check the returned report; stop on NEEDS-WORK.

4) Aggregate report

List every produced file, its artifact, and the wiring (which entity_id references which). Relay each dispatched skill's CONFORMANT / NEEDS-WORK report verbatim — do not re-judge them. Do not deploy.

Boundaries

  • Single-artifact generation + spec conformance → the owning authoring skill
  • Custom integration → ha-integration-scaffold (this skill only recognizes and points)
  • Deploy to live HA → out of scope

Resumability

Re-invoking this skill with the same requirement resumes per spec/claude/resumable-work/: the checkpoint under .resume/<skill-name>/ records the approved plan and per-step dispatch status, so an interrupted orchestration continues at the first incomplete step instead of re-planning from scratch.

Keep looking

Skills are one crate of 608,666. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.