Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add nolte/claude-home-assistant --skill ha-esphome-package-authorgit clone --depth 1 https://github.com/nolte/claude-home-assistantWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/nolte/claude-home-assistant/ha-esphome-package-author)<a href="https://agentmods.dev/skills/nolte/claude-home-assistant/ha-esphome-package-author"><img src="https://agentmods.dev/badge/skills/nolte/claude-home-assistant/ha-esphome-package-author/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/nolte/claude-home-assistant/ha-esphome-package-author"><img src="https://agentmods.dev/badge/skills/nolte/claude-home-assistant/ha-esphome-package-author.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00203 | $0.02166 |
| Opus 5 | $0.00102 | $0.01083 |
| Sonnet 5 | $0.00041 | $0.00433 |
| Haiku 4.5 | $0.00020 | $0.00217 |
Grade A, and why
ha-esphome-package-author scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
HA ESPHome Package Author
Spec: spec/claude/ha-esphome-package-author/en.md (EN canonical) / spec/claude/ha-esphome-package-author/de.md (DE translation). Grounding spec: spec/ha/esphome-project-structure/en.md §Package architecture, §Parameterisation, §Deviating without forking, §Remote packages.
Owns one shared package per invocation: how it is cut, what its parameter interface is, and how a deviating device consumes it without forking it.
Why this is a skill, not an agent
- Mid-flow approval is the contract (decisive): where the concern boundary runs, which variables become the package's public interface, and whether a deviation is an
!extendor a new parameter are judgement calls the operator confirms — and each one is a breaking change for existing consumers if it is wrong. - Quick, targeted change in the current context: one file under
common/plus the consuming device files, iterated in conversation. - Counter-dimension considered: reading every consumer of the package before writing is a bounded fan-out that could be isolated (agent bias), but the resulting interface decision must be discussed, not reported; skill wins.
When this skill activates
The user wants shared ESPHome configuration to become or stay one well-cut package — "das gehört ins gemeinsame Package", "make the board package take a parameter for the update interval", "box-02 needs the base package without captive_portal".
When NOT to activate
- the repository tree, the initial base/board packages →
ha-esphome-fleet-scaffold - a block that belongs to exactly one device →
ha-esphome-config-augment - a new device file →
ha-esphome-config-scaffold - a read-only verdict across the whole package architecture →
ha-esphome-fleet-reviewer - compiling, flashing, or rolling out → the ESPHome toolchain / operator
Hard rules
- Read the grounding spec, the package, and every consumer first. A package edit reaches every device that includes it; generating one from memory is how a fleet breaks silently.
- One concern per package. A board package, a base package, a feature package — never a grab bag, and never a near-duplicate that differs from an existing package in one key.
- Two occurrences make a parameter. One deviation is an exception resolved with
!extend/!removeon the component's config id; the second identical override is promoted into the package as a defaulted variable.!extendnever targets a package key — package keys are documentation only. - Every variable a package reads that not every consumer sets carries a
defaults:entry. Adding a variable without a default is a breaking change for existing consumers, and this skill states that in the report. id:on anything overridable. A component a device might extend or remove gets an explicitid:; otherwise the merge concatenates and the device ends up with two components.- State the merge consequences before writing. Dictionaries merge key-by-key, component lists merge by id, other lists concatenate, and every other value is replaced by the later one — the last clause is the one that silently lets a device scalar win.
- Credentials stay
!env_var, and no!secretenters a package that is or could become remote.!env_varis undocumented API and cannot interpolate substitutions in its name; new variables get documented in the same run. - Remote packages are pinned or vendored. A tag or commit, never a moving branch without a deliberate, documented
refresh; prefer vendoring intocommon/. - Keep nesting shallow (device → board → base) and pass parameters in through substitutions rather than letting a package reach back into a consumer's secrets.
- One package per run. No fleet-wide refactoring batches; validation is reported, never silently skipped.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 103 lines · 203 tokens per session scan A ca7aea2e5c0e
ha-esphome-package-author is a skill published in the GitHub repository nolte/claude-home-assistant (1 stars, last pushed 1mo ago), licensed MIT. It adds 203 tokens to every session and 2,166 once invoked, about $0.0010 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
embedded-iot
Embedded systems firmware, microcontrollers (ESP32, STM32, Arduino, Raspberry Pi), RTOS (FreeRTOS, Zephyr), IoT protocols (MQTT, CoAP, BLE), bare-metal C/C++, and hardware peripheral interfaces (I2C, SPI, UART, GPIO). Use when developing firmware, working with microcontrollers, or building IoT devices.
spark-environment-setup
Set up a working ML training/inference environment on NVIDIA DGX Spark (GB10, aarch64, CUDA 13). Use when installing PyTorch/Unsloth/TRL/vLLM on DGX Spark, hitting libcudart or wheel-ABI errors on aarch64, or choosing between NGC containers and bare pip installs.
spark-training-gotchas
Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.
security-compliance
Guides security professionals in implementing defense-in-depth security architectures, achieving compliance with industry frameworks (SOC2, ISO27001, GDPR, HIPAA), conducting threat modeling and risk assessments, managing security operations and incident response, and embedding security throughout the SDLC.
stride-analysis-patterns
Apply STRIDE methodology to systematically identify threats. Use when analyzing system security, conducting threat modeling sessions, or creating security documentation.
review-loop
Run the adversarial verification loop — implement, then hand the change to a fresh checker that did not write it, fix what it finds, and re-dispatch until APPROVE. Use before claiming any behavioural change is done, and on requests like "review loop", "adversarial review", "independent review", "get this verified"…