/ha-analyze-patterns

name: ha-analyze-patterns
description: Analyze Home Assistant history data and entity patterns to identify automation opportunities, unused devices, and energy anomalies. Use periodically or when looking for optimization opportunities.
allowed-tools:
  - Bash
  - Read
  - Write
  - Glob
  - Grep
  - mcp__homeassistant__GetLiveContext
  - mcp__homeassistant__homeassistant__GetLiveContext
  - mcp__homeassistant__GetDateTime
  - mcp__homeassistant__llm__GetDateTime

HA Pattern Analysis

Steps

1. **Load existing analysis**: Read `.hermit/raw/snapshot-ha-pattern-analysis-latest.json` if it exists. 2. **Get live context**: Call `GetLiveContext` and `GetDateTime` via MCP for current state. 3. **Fetch history**: Run `${CLAUDE_PLUGIN_ROOT}/bin/ha-agent-lab ha fetch-history --window-days 7`. When it prints `{"error"}` (exit 1), pipe one line into `.hermit/bin/hermitd-run task note .hermit <id>` only inside an open record's turn (otherwise skip the note) (`history fetch failed: <error>`) and continue: the skill proceeds using Phase 1 silence findings only. Do not surface the fetch failure to scheduled-check stdout. 4. **Read normalized inventory**: Read `.hermit/raw/snapshot-ha-normalized-latest.json`. Extract both `entity_index` and `silence_summary` (added by every context refresh). Also read `.hermit/raw/snapshot-ha-history-7d-latest.json` if present. 5. **Analyze patterns from silence_summary and history**:

From `silence_summary.dead_automations` (enabled automations not fired in 30+ days):

• Each entry is an actionable Reliability issue: `"automation.X has not fired in N days (enabled)"`.
• `never_fired: true` entries get special phrasing: `"automation.X has never fired (enabled but never triggered)"`.

From `silence_summary.silent_event_sensors` (motion/door/window sensors silent for 7+ days):

• Each entry is an actionable Reliability issue: `"binary_sensor.X (<device_class>) — no events in N days"`.

From `silence_summary.long_unavailable` (individual entities unavailable 7+ days, domain not already degraded):

• Each entry is a Reliability issue: `"entity.X unavailable for N days"`.

From `silence_summary.inactive_candidates_by_domain` (lights/switches/covers/climates unchanged for 7+ days):

• Surface in the Markdown patterns artifact only. Do **not** include in the stdout findings block — these are informational, not actionable without corroboration.

From `silence_summary.suppressed_entity_domains`:

• Note in Markdown artifact that these domains are already covered by ha-integration-health.

From `snapshot-ha-history-7d-latest.json` (when present):

• For each entry in `time_patterns`: surface under `Automation opportunities:` as a schedule-based candidate. Example: `"light.kitchen — 8/12 events at 07:00 UTC, consider time-based automation"`.
• For each `entity_aggregates` entry where `event_count == 0` AND the entity also appears in `silence_summary.inactive_candidates_by_domain`: promote to actionable `Reliability issues:` **only** when the entity's `last_changed` timestamp (from entity_index) is older than 30 days. The 7-day Phase 1 informational threshold gets a much higher bar here because history corroborates it.
• Do not derive automation execution from history entries for `automation.*` entities — `silence_summary.dead_automations` is authoritative for that.

6. **Write findings**: Save pattern data to `.hermit/raw/snapshot-ha-pattern-analysis-<YYYY-MM-DD>.json` and update `snapshot-ha-pattern-analysis-latest.json`. Write the curated Markdown summary (when non-trivial findings exist) to `.hermit/raw/patterns-<YYYY-MM-DD>.md` with frontmatter `type: analysis`, `title: "HA Pattern Analysis: <YYYY-MM-DD>"`, `created: <ISO8601>`, `task: <T-... for the open record in this turn; omit this field otherwise>`, `tags: [ha-patterns, analysis]`. Also write `patterns-latest.md` pointing to the same content. The Markdown body should include the richer breakdown: `inactive_candidates_by_domain` summary table, `suppressed_entity_domains` note, time-pattern table from history, and `state_durations` summary for `climate.*` entities (showing heating/cooling hours over the 7d window). 7. **Update memory**: Update your auto memory with key findings from this analysis. 8. **Emit summary for reflect --check-id ha-patterns --check hermitd-homeassistant:ha-analyze-patterns**: Always output a plain-text findings block to stdout. reflect --check-id ha-patterns --check hermitd-homeassistant:ha-analyze-patterns routes actionable items through the proposal pipeline. The stdout shape is fixed — do not add new top-level sections:

   ha-analyze-patterns findings — <date>
   Automation opportunities: <N>
   - <opportunity 1>
   Reliability issues: <N>
   - <dead automation or silent sensor or long-unavailable entity>
   Waste patterns: <N>
   - <waste 1>
   No action needed: <list anything normal or already automated>

Dead automations and silent event sensors fold under `Reliability issues:`. If there are zero findings across all categories, output: `ha-analyze-patterns findings — <date>\nNo actionable findings.`

9. **Propose automations**: If clear patterns emerge, suggest them. For complex ones, delegate to `@hermitd-homeassistant:ha-automation-builder`.

What to Look For

• **Dead automations**: enabled automations with no recent triggers — often broken silently after an HA upgrade
• **Silent event sensors**: motion/door/window sensors that haven't fired — may be stuck, dead, or unused
• **Long-unavailable entities**: individual entities down for 7+ days (domain not covered by integration-health)
• **Time patterns**: lights/devices that follow daily schedules; repeated manual actions the operator performs at the same time of day — both are candidates for time-based automations
• **Correlation patterns**: events that always happen together (candidates for grouped automations); manual action sequences that could be a single trigger

-