Architecture review — Home Assistant Homelab
SkillFiles & storageLets your agent review a codebase for design problems like tight coupling and missing boundaries.
Use Architecture review — Home Assistant Homelab in Claude, ChatGPT or Ahel Desktop
Free. Sign in, add Architecture review — Home Assistant Homelab and connect your AI. About a minute.
Also: Claude Code · Cursor · Codex
Then ask your AI: use the Architecture review skill
Details
Instructions available. Your AI can read the instructions. Execution depends on the setup they require.
Account requirements not reviewed. Check the skill instructions before use; Ahel provides instructions and does not run this skill.
No other account needed.
Add Ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.
About this skill
Explore packages/, layout, and .agents/ for architectural friction (coupling, shallow packages, missing boundaries) and file or recommend agent issues under docs/issues/. Use when improving config shape, before large refactors, or when find-work should surface design-debt candidates.
What this skill tells your AI
The instructions your AI receives, as published by apocrathia/home-assistant-config in .agents/skills/architecture-review/SKILL.md and read by Ahel’s review.
Read-only exploration that ends in candidates the operator can pick, then
optional filing under docs/issues/ (kind: architecture) via
file-issue. Git is operator-owned
(operator-owned-git.md). No worktrees
(worktrees.md).
Human Local Todo stays human — do not write ICS. Prefer docs issues for agent-owned architecture debt.
Deliverable
A numbered candidate list. For each pick the operator cares about: Problem, Proposed direction, Feedback loop, and either:
- File
docs/issues/<slug>.mdfrom_template.md(kind: architecture), or - Leave as chat-only if the operator declines filing
Workflow
- [ ] 1. Orient: AGENTS.md, constraints + traps, packages layout
- [ ] 2. Explore (read-only): packages/, esphome/, docs/, .agents/
- [ ] 3. Present numbered candidates
- [ ] 4. Operator picks (or asks for a deeper sketch on rank-1)
- [ ] 5. For each pick: problem space + package sketch (chat); no code edits
- [ ] 6. On operator OK: file-issue for each kept candidate
What to look for
- Shallow packages / cross-area coupling
- Logic bouncing across many files for one routine
- Secrets risking hardcoding / wrong package home
- Automations without a clear validation path
- ESPHome / packages / docs drift
.agents/routing that points at missing paths- Untested assumptions about live entities
Candidate content (per pick)
- Problem — packages/modules involved
- Proposed direction —
packages/areas|functions|integrations|projects|routines|system|toys/ - Feedback loop —
config-validate, reload, live entity check, automation trace - Suggested next — alignment / plan / implement-change / file-issue
Orient paths
| Path | Why |
|---|---|
AGENTS.md | Router |
packages/ | Modular YAML |
esphome/ | Device definitions |
.agents/context/ | Constraints, traps, nomenclature |
docs/issues/ | Agent issue ledger |
Do not
- Edit packages, ESPHome, or
.agents/in this skill (exploration only) - Write Local Todo ICS or
.shopping_list.json - Create GitHub issues unless the operator asks
- File vague "refactor X" without a named feedback loop
- Commit, push, open PRs, or create worktrees
Signals
- GitHub stars
- 179
- Forks
- 32
- Last commit
- Sep 2026
Advanced
- Item type
- skill
- Key
architecture-review- Source
- github.com/apocrathia/home-assistant-config
github.com/apocrathia/home-assistant-config