[libretiny] Report the previous boot's crash on RTL8720C (#21648)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: J. Nick Koston <nick@koston.org>

.agents/skills/automations/SKILL.md

@@ -0,0 +1,215 @@
+---
+name: automations
+description: Write ESPHome automation building blocks - triggers (on_* keys), actions and conditions. Use when adding or changing an on_* trigger, an action such as my_component.do_something, a condition, a callback registration method (add_on_*_callback), or code that calls build_callback_automation, build_automation, register_*_action or register_*_condition.
+---
+
+# Automations: triggers, actions and conditions
+
+Automations have three building blocks: **Triggers** (fire when something happens), **Actions** (do
+something), and **Conditions** (check if something is true). The helpers named here are documented in
+their docstrings in `esphome/automation.py`.
+
+## Triggers: callback method (preferred)
+
+Use `build_callback_automation()` for simple triggers. This eliminates the need for a C++ Trigger class by
+using a lightweight pointer-sized forwarder struct registered directly as a callback. No `CONF_TRIGGER_ID`
+in the schema.
+
+**Python:**
+```python
+from esphome import automation
+
+CONFIG_SCHEMA = cv.Schema(
+    {
+        cv.GenerateID(): cv.declare_id(MyComponent),
+        cv.Optional(CONF_ON_STATE): automation.validate_automation({}),
+    }
+).extend(cv.COMPONENT_SCHEMA)
+
+
+async def to_code(config):
+    var = cg.new_Pvariable(config[CONF_ID])
+    await cg.register_component(var, config)
+    for conf in config.get(CONF_ON_STATE, []):
+        await automation.build_callback_automation(
+            var, "add_on_state_callback", [(bool, "x")], conf
+        )
+```
+
+`build_callback_automation` arguments: `parent`, `callback_method` (C++ method name), `args` (template args
+as `[(type, name)]` tuples), `config`, and optional `forwarder` (defaults to `TriggerForwarder<Ts...>`).
+
+For boolean filtering (e.g. `on_press`/`on_release`), use built-in forwarders with `args=[]`:
+```python
+for conf_key, forwarder in (
+    (CONF_ON_PRESS, automation.TriggerOnTrueForwarder),
+    (CONF_ON_RELEASE, automation.TriggerOnFalseForwarder),
+):
+    for conf in config.get(conf_key, []):
+        await automation.build_callback_automation(
+            var, "add_on_state_callback", [], conf, forwarder=forwarder
+        )
+```
+
+When the callback's parameters differ from the automation's arguments, or the trigger should only fire for
+some values, pass `params`, `forward` and `when`; the helper then generates a capture-less lambda (stored
+inline, nothing allocated). Name the parent with `automation.parent_ref(var)` and build `forward` from
+`MockObj` calls, not f-strings of C++:
+```python
+parent = automation.parent_ref(var)
+index = cg.RawExpression("index")
+await automation.build_callback_automation(
+    var,
+    "add_on_state_callback",
+    [(cg.StringRef, "x"), (cg.size_t, "i")],
+    conf,
+    params=[(cg.size_t, "index")],
+    forward=[cg.StringRef(parent.option_at(index)), index],
+)
+```
+`build_parent_callback_automation`, `build_trigger_callback`, `build_callback_automations` and
+`build_trigger_automations` cover the other shapes; see their docstrings in `esphome/automation.py`.
+
+**C++: no trigger class needed.** The callback registration method must be templatized to accept both
+`std::function` and lightweight forwarder structs (which avoid heap allocation):
+```cpp
+class MyComponent : public Component {
+ public:
+  // Must be a template -- accepts both std::function and pointer-sized forwarder structs
+  template<typename F> void add_on_state_callback(F &&callback) {
+    this->state_callback_.add(std::forward<F>(callback));
+  }
+ protected:
+  // Use CallbackManager when callbacks are always registered (e.g. core components)
+  CallbackManager<void(bool)> state_callback_;
+  // Use LazyCallbackManager when callbacks are often not registered -- saves 8 bytes
+  // (nullptr vs empty std::vector) per instance when no callbacks are added
+  // LazyCallbackManager<void(bool)> state_callback_;
+};
+```
+
+Never type a registration method's parameter as `std::function`; that forces a heap allocation for every
+forwarder struct:
+```cpp
+// Bad
+void add_on_state_callback(std::function<void(bool)> &&callback) {
+  this->state_callback_.add(std::move(callback));
+}
+```
+
+| Type | Idle overhead (32-bit) | When to use |
+|------|----------------------|-------------|
+| `CallbackManager<void(Ts...)>` | 12 bytes (empty `std::vector`) | Callbacks are always or almost always registered |
+| `LazyCallbackManager<void(Ts...)>` | 4 bytes (`nullptr`) | Callbacks are often not registered (common case) |
+
+Both live in `esphome/core/helpers.h` and accept `std::function`, lambdas and forwarder structs through
+their templatized `add()` method. `LazyCallbackManager` is a drop-in replacement for `CallbackManager` that defers allocation until the first
+callback is added. Prefer it for entity-level callbacks where most instances have no subscribers.
+
+## Triggers: Trigger class method
+
+Use `build_automation()` with a `Trigger<Ts...>` subclass only when the forwarder needs **mutable state
+beyond a single `Automation*` pointer** (e.g. edge detection tracking previous state, timing logic).
+
+**Python:**
+```python
+TurnOnTrigger = my_ns.class_("TurnOnTrigger", automation.Trigger.template())
+
+CONFIG_SCHEMA = cv.Schema(
+    {
+        cv.Optional(CONF_ON_TURN_ON): automation.validate_automation(
+            {cv.GenerateID(CONF_TRIGGER_ID): cv.declare_id(TurnOnTrigger)}
+        ),
+    }
+)
+
+
+async def to_code(config):
+    for conf in config.get(CONF_ON_TURN_ON, []):
+        trigger = cg.new_Pvariable(conf[CONF_TRIGGER_ID], var)
+        await automation.build_automation(trigger, [], conf)
+```
+
+**C++:**
+```cpp
+class TurnOnTrigger : public Trigger<> {
+ public:
+  explicit TurnOnTrigger(MyComponent *parent) : last_on_{false} {
+    parent->add_on_state_callback([this](bool state) {
+      if (state && !this->last_on_)
+        this->trigger();
+      this->last_on_ = state;
+    });
+  }
+ protected:
+  bool last_on_;
+};
+```
+
+## Actions
+
+```cpp
+template<typename... Ts> class MyAction : public Action<Ts...> {
+ public:
+  explicit MyAction(MyComponent *parent) : parent_(parent) {}
+  void play(const Ts &...) override { this->parent_->do_something(); }
+ protected:
+  MyComponent *parent_;
+};
+```
+Register it without writing a builder:
+```python
+automation.register_simple_action(
+    "my_component.do_something", MyAction, schema, synchronous=True
+)
+```
+The constructor receives the object named by `config[CONF_ID]`. Use `register_bare_action` for a
+no-argument constructor, `register_parented_action` for a class deriving from `Parented<T>`, and the
+`@automation.register_action(...)` decorator only when the builder must also set fields.
+
+Use `synchronous=True` for actions that run to completion inside `play()` without deferring. Use
+`synchronous=False` if the action may suspend/defer execution (e.g. `delay`, `wait_until`, `script.wait`)
+or store trigger arguments for later use.
+
+**Actions that only forward templatable values to their parent need no C++ class.** Register them with
+`register_apply_action`; do not write a `TEMPLATABLE_VALUE` class or a builder for this shape.
+```python
+automation.register_apply_action(
+    "my_component.set_gains",
+    schema,
+    automation.ApplyField(CONF_KP, "set_kp", cg.float_),
+    automation.ApplyField(CONF_KI, "set_ki", cg.float_),
+)
+```
+The `ApplyField`, `ApplyCall` and `register_apply_action` docstrings in `esphome/automation.py` cover the
+rest; `cover.control` and `cover.template.publish` are in-tree examples. An action whose input is a byte
+payload passes `automation.ApplyBytes(CONF_DATA, "send({})")` in the same call; a constant payload stays in
+flash and a lambda's vector is forwarded as data and size. `TEMPLATABLE_VALUE` with `cg.templatable` stays
+for actions whose `play()` has real logic beyond forwarding values.
+
+## Conditions
+
+```cpp
+template<typename... Ts> class MyCondition : public Condition<Ts...> {
+ public:
+  explicit MyCondition(MyComponent *parent) : parent_(parent) {}
+  bool check(const Ts &...) override { return this->parent_->is_active(); }
+ protected:
+  MyComponent *parent_;
+};
+```
+Register with `automation.register_simple_condition("my_component.is_active", MyCondition, schema)`;
+`register_bare_condition`, `register_parented_condition` and the decorator follow the action rules.
+
+**Conditions that only test their parent need no C++ class either.** Register them with
+`register_apply_condition`; the expression is applied to the parent, and an `ApplyCall` compares against
+config values.
+```python
+automation.register_apply_condition("my_component.is_active", schema, "is_active()")
+automation.register_apply_condition(
+    "my_component.state_is",
+    schema,
+    automation.ApplyCall("state == {}", ((CONF_STATE, cg.bool_),)),
+)
+```
+`cover.is_open`, `rtttl.is_playing` and `component.is_idle` are in-tree examples.

.agents/skills/breaking-changes/SKILL.md

@@ -0,0 +1,68 @@
+---
+name: breaking-changes
+description: Decide whether an ESPHome change breaks public API and handle it - what counts as public C++ and Python API, the deprecation window, ESPDEPRECATED for C++, cv.rename_key and deprecation warnings for config keys, and the migration notes a PR needs. Use when removing, renaming or changing the signature of a config option, a public C++ member or a core Python helper.
+---
+
+# Public API and breaking changes
+
+Full policy: https://developers.esphome.io/contributing/code/#public-api-and-breaking-changes
+
+## What is public
+
+- **C++ in components:** only features documented at [esphome.io](https://esphome.io) are public API.
+  Undocumented `public` members are internal.
+- **C++ core and base classes** (`esphome/core/`, `Component`, `Sensor`, etc.): all `public` members are
+  public API.
+- **Components with global accessors** (`global_api_server`, etc.): all `public` members are public API
+  (except config setters).
+- **Python:** all documented configuration options are public API. Python code in `esphome/core/` that
+  existing core components use is stable API. Other Python code is internal unless explicitly documented
+  for external component use.
+- Changing the signature of a lambda that code generation injects is not a breaking change.
+
+## Policy
+
+- Aim for a **6-month deprecation window** when possible.
+- Clean breaks are allowed for signature changes, deep refactorings and resource constraints.
+- The migration path goes in the PR description (it generates the release notes).
+- A blog post is required for core/base class changes or significant architectural changes.
+
+Checklist:
+- [ ] Clear justification (RAM/flash savings, architectural improvement)
+- [ ] Explored non-breaking alternatives
+- [ ] Added deprecation warnings if possible (`ESPDEPRECATED` for C++)
+- [ ] Documented migration path in the PR description with before/after examples
+- [ ] Updated all internal usage and esphome.io
+- [ ] Tested backward compatibility during the deprecation period
+
+## Deprecating C++
+
+```cpp
+// Remove before 2026.6.0
+ESPDEPRECATED("Use new_method() instead. Removed in 2026.6.0", "2025.12.0")
+void old_method() { this->new_method(); }
+```
+
+## Deprecating config keys
+
+For a renamed key, use the shared `cv.rename_key` validator with `removed_in` (and `component` for
+context); it warns and migrates automatically:
+```python
+CONFIG_SCHEMA = cv.All(
+    cv.rename_key(
+        CONF_OLD_KEY, CONF_NEW_KEY, removed_in="2026.6.0", component="my_component"
+    ),
+    cv.Schema({...}),
+)
+```
+For other deprecations, warn manually during validation:
+```python
+# Remove before 2026.6.0
+if CONF_OLD_KEY in config:
+    _LOGGER.warning(
+        f"'{CONF_OLD_KEY}' deprecated, use '{CONF_NEW_KEY}'. Removed in 2026.6.0"
+    )
+    config[CONF_NEW_KEY] = config.pop(CONF_OLD_KEY)  # Auto-migrate
+```
+Cover the old syntax with a `validate.*.yaml` test (see the `component-tests` skill), and keep the language
+schema correct (see the `config-validation` skill).

.agents/skills/code-review/SKILL.md

@@ -9,11 +9,13 @@ ESPHome parses YAML into C++ firmware for memory-constrained microcontrollers
 (ESP32, ESP8266, RP2040, LibreTiny). Review changes with that in mind: RAM and
 flash are scarce, and code runs unattended for months.
 
-`AGENTS.md` in the repository root is the full contributor guide and the
+`AGENTS.md` in the repository root is the contributor guide, and the skills it
-authority when it disagrees with this summary. The developer documentation at
+lists (`component-codegen`, `config-validation`, `automations`,
-https://developers.esphome.io explains the component lifecycle and the reasoning
+`embedded-memory`, `component-tests`, `breaking-changes`) hold the detail. They
-behind these rules. This skill lists the concrete things worth flagging in a
+are the authority when they disagree with this summary. The developer
-review; read `AGENTS.md` for the detail behind any item.
+documentation at https://developers.esphome.io explains the component lifecycle
+and the reasoning behind these rules. This skill lists the concrete things worth
+flagging in a review; read the matching skill for the detail behind any item.
 
 Only raise findings that the diff actually introduces or changes. Do not ask for
 drive-by cleanup of pre-existing code the PR did not touch.
@@ -76,6 +78,10 @@ performance nit, because it fragments a small shared heap. Flag:
   `if (blah := config.get(CONF_BLAH)) is not None:`.
 - Reuse existing validators from `config_validation.py` (`cv.rename_key`,
   `cv.has_exactly_one_key`, etc.) via `cv.All(...)` instead of hand-rolling.
+- A new or changed validator function that wraps, chooses or builds a schema
+  must be decorated with `@schema_extractor(...)` and handle `SCHEMA_EXTRACT`,
+  or `script/build_language_schema.py` silently drops the keys behind it (see
+  the `config-validation` skill).
 - `esphome/const.py` is frozen: no new `CONF_` constants there. Define them in
   the component's own `.py`, or in `esphome/components/const/__init__.py` when
   shared. The same constant defined in three or more component files fails CI.

.agents/skills/component-codegen/SKILL.md

@@ -0,0 +1,240 @@
+---
+name: component-codegen
+description: Write the Python side of an ESPHome component - __init__.py and platform files (sensor.py, switch.py, ...), DOMAIN and other metadata, to_code, hub child entities with sub_sensors and friends, pins and referenced components, hub-forwarding entities (ApplyButton, ApplySwitch), uart/i2c/spi constructor helpers, and CORE.data state. Use when creating a component or changing its code generation.
+---
+
+# Component code generation
+
+Read https://developers.esphome.io/architecture/components/ for the component lifecycle and how a
+component is registered. Schema rules and schema extraction are in the `config-validation` skill;
+automations (`on_*`, actions, conditions) in the `automations` skill.
+
+## Files
+
+```
+components/[component_name]/
+├── __init__.py          # Component configuration schema and code generation
+├── [component].h        # C++ header file (if needed)
+├── [component].cpp      # C++ implementation (if needed)
+└── [platform]/          # Platform-specific implementations
+    ├── __init__.py      # Platform-specific configuration
+    ├── [platform].h     # Platform C++ header
+    └── [platform].cpp   # Platform C++ implementation
+```
+
+## Metadata
+
+- `DOMAIN`: required in every component's `__init__.py`, set to the component's own name as a plain string
+  (e.g. `DOMAIN = "my_component"`). Other code refers to the component through it, e.g.
+  `from esphome.components.my_component import DOMAIN`. Keep it in alphabetical order with the other
+  metadata constants. CI (`lint_component_domain`) fails without it.
+- `DEPENDENCIES`: list of required components.
+- `AUTO_LOAD`: components to automatically load. Not a way to put entities under the component's own key;
+  see "Entities of a hub" below.
+- `CONFLICTS_WITH`: incompatible components.
+- `CODEOWNERS`: GitHub usernames responsible for maintenance.
+- `MULTI_CONF`: whether multiple instances are allowed.
+
+## Basic pattern
+
+```python
+import esphome.codegen as cg
+import esphome.config_validation as cv
+from esphome.const import CONF_KEY, CONF_ID
+
+DOMAIN = "my_component"
+CONF_PARAM = "param"  # A constant that does not yet exist in esphome/const.py
+
+my_component_ns = cg.esphome_ns.namespace("my_component")
+MyComponent = my_component_ns.class_("MyComponent", cg.Component)
+
+CONFIG_SCHEMA = cv.Schema(
+    {
+        cv.GenerateID(): cv.declare_id(MyComponent),
+        cv.Required(CONF_KEY): cv.string,
+        cv.Optional(CONF_PARAM, default=42): cv.int_,
+    }
+).extend(cv.COMPONENT_SCHEMA)
+
+
+async def to_code(config):
+    var = cg.new_Pvariable(config[CONF_ID])
+    await cg.register_component(var, config)
+    cg.add(var.set_key(config[CONF_KEY]))
+    cg.add(var.set_param(config[CONF_PARAM]))
+```
+
+```cpp
+namespace esphome::my_component {
+
+class MyComponent : public Component {
+ public:
+  void setup() override;
+  void loop() override;
+  void dump_config() override;
+
+  void set_key(const char *key) { this->key_ = StringRef(key); }
+  void set_param(int param) { this->param_ = param; }
+
+ protected:
+  StringRef key_;
+  int param_{0};
+};
+
+}  // namespace esphome::my_component
+```
+
+When the setter takes the value unchanged, use `cg.add_optional(var.set_blah, config.get(CONF_BLAH))`; it
+is not awaited, and it skips only `None`, so `0` and `False` are still set. Required, invariant
+dependencies are constructor arguments to `cg.new_Pvariable()`, not setters.
+
+## Entity platforms
+
+**Sensor:**
+```python
+from esphome.components import sensor
+
+CONFIG_SCHEMA = sensor.sensor_schema(MySensor).extend(
+    cv.polling_component_schema("60s")
+)
+
+
+async def to_code(config):
+    var = await sensor.new_sensor(config)
+    await cg.register_component(var, config)
+```
+
+**Binary sensor:**
+```python
+from esphome.components import binary_sensor
+
+CONFIG_SCHEMA = binary_sensor.binary_sensor_schema().extend({...})
+
+
+async def to_code(config):
+    var = await binary_sensor.new_binary_sensor(config)
+```
+
+**Switch:**
+```python
+from esphome.components import switch
+
+CONFIG_SCHEMA = switch.switch_schema().extend({...})
+
+
+async def to_code(config):
+    var = await switch.new_switch(config)
+```
+
+## Entities of a hub
+
+A new component's entities go in platform files (`sensor.py`, `binary_sensor.py`, ...) that take the hub's
+id, never under the hub's own key with their platforms added to `AUTO_LOAD` to make that work. Auto loading
+a platform because the C++ uses its type is fine. See
+https://developers.esphome.io/architecture/components/#exposing-multiple-entities.
+```yaml
+# Bad - entities under the hub's key
+my_hub:
+  connected:
+    name: Connected
+
+# Good - the hub, then one platform entry that holds its entities
+my_hub:
+  id: hub1
+binary_sensor:
+  - platform: my_hub
+    my_hub_id: hub1  # optional with a single hub
+    connected:
+      name: Connected
+```
+
+**Optional child entities of a hub:** bind the config once with `sensor.sub_sensors(config)` (or
+`sub_binary_sensors`, `sub_text_sensors`, `sub_buttons`, `sub_switches`, `sub_numbers`, `sub_selects` in
+their domains), adding `parent=hub` for entities that derive from `Parented<T>`, then make one call per
+key, even when there is only one. A call creates the entity only when its key is configured, passes it to
+the setter and returns it (or `None`); extra arguments such as `min_value` or `options` go on the call.
+Always name the setter explicitly on the object that owns it, never with `getattr` and an f-string, and
+keep that variable short (`var` for the component itself, `hub` for one fetched with `cg.get_variable`) so
+the calls fit on one line. Loops whose setter also takes an index, such as `set_gate_threshold(x, n)`, stay
+as they are.
+```python
+# my_hub/sensor.py
+async def to_code(config):
+    hub = await cg.get_variable(config[CONF_MY_HUB_ID])
+    sensors = sensor.sub_sensors(config)
+    await sensors(CONF_TEMPERATURE, hub.set_temperature_sensor)
+    await sensors(CONF_HUMIDITY, hub.set_humidity_sensor)
+
+
+# my_hub/button.py
+async def to_code(config):
+    hub = await cg.get_variable(config[CONF_MY_HUB_ID])
+    buttons = button.sub_buttons(config, parent=hub)
+    await buttons(CONF_RESTART, hub.set_restart_button)
+```
+
+**Hub forwarding entities need no C++ class:** a button, switch, number or select that only calls its hub
+uses `button.ApplyButton` (`switch.ApplySwitch`, `number.ApplyNumber`, `select.ApplySelect`) in its schema
+and `apply=` on the `sub_*` call. The body is a call on the hub with `entity` and `state`, `value` or
+`index`/`option` in scope; on a switch, number or select `publish="before"` or `"after"` publishes the
+state. `publish="momentary"` is only for legacy action switches (restart, factory reset): it publishes off
+and runs the body only when turned on. Never use it for a new entity; an action is a button.
+```python
+switches = switch.sub_switches(config, parent=hub)
+await switches(CONF_BLUETOOTH, None, apply="set_bluetooth(state)", publish="after")
+```
+Restore an owned or standalone switch in `setup()` with `sw->restore()`, or `restore(default)` to still
+drive a state when the restore mode is disabled. A hub that applies its switches' restored states itself
+reads `sw->get_initial_write_state()`, which already applies `inverted`.
+
+**Entities that each watch a register field:** one class taking register, shift, mask and match from a
+Python tuple table, in typed `StaticVector`s sized by `cg.slot_counter`; not a template per bit, not a
+virtual listener base (a second vptr per entity).
+
+## Other helpers
+
+- **Pins and referenced components:** set them with `await cg.add_pin(var.set_cs_pin, config[CONF_CS_PIN])`
+  and `await cg.add_variable(var.set_bus, config[CONF_BUS_ID])`; both skip a `None` value, so pass
+  `config.get(KEY)` for an optional key, and return what they created. Use `cg.register_parented` for
+  `set_parent`.
+- `uart.new_uart_component(config, *ctor_args)` (and `i2c.new_i2c_component`, `spi.new_spi_component`,
+  `modbus.new_modbus_client_component`, `ble_client.new_ble_node_component`) constructs, registers and
+  attaches in one call.
+- `display.add_writer(var, config)` sets the lambda.
+- `MyEnum.options("on", prefix="MY_ENUM_")` builds an enum dict.
+- `button.restart_button_schema(cls)` (and `factory_reset_`, `query_`) are button presets.
+- `fv.get_declaring_config(id)` finds a hub's config in final validation.
+
+## State during code generation
+
+Use `CORE.data` for component state that needs to persist during configuration generation, namespaced
+under the component's `DOMAIN` and held in a `@dataclass`. Avoid module-level mutable globals: they
+persist between compilation runs if the host process (e.g. device-builder) doesn't fork/exec, whereas
+`CORE.data` is cleared between runs.
+
+```python
+from dataclasses import dataclass, field
+from esphome.core import CORE
+
+DOMAIN = "my_component"
+
+
+@dataclass
+class MyComponentData:
+    feature_enabled: bool = False
+    item_count: int = 0
+    items: list[str] = field(default_factory=list)
+
+
+def _get_data() -> MyComponentData:
+    if DOMAIN not in CORE.data:
+        CORE.data[DOMAIN] = MyComponentData()
+    return CORE.data[DOMAIN]
+
+
+def request_feature() -> None:
+    _get_data().feature_enabled = True
+```
+
+Don't use flat keys such as `CORE.data["my_component_feature"] = True`. Some components use `TypedDict`
+instead of a dataclass; both are acceptable.

.agents/skills/component-tests/SKILL.md

@@ -0,0 +1,101 @@
+---
+name: component-tests
+description: Write and run tests for ESPHome components - YAML build tests (test.*.yaml, validate.*.yaml, common.yaml) under tests/components, shared bus packages and CI test grouping, C++ gtest unit tests, and host integration tests including the uart_mock fixture. Use when adding or changing files under tests/, or when a code change needs test coverage.
+---
+
+# Component tests
+
+## C++ unit tests
+
+gtests live in `tests/components/<component>/*.cpp` and run with `script/cpp_unit_test.py <component>`; see
+`tests/components/README.md`. Set the loop clock in a test with `LoopBlockingGuard{nullptr, nullptr,
+now_ms}`; the runner resets it before every test.
+
+## Integration tests
+
+`tests/integration/` runs real firmware on the host. Serial components prove timeouts and recovery with the
+`uart_mock` fixture (`tests/integration/test_uart_mock_*.py`). Its `responses` match on content every time
+they see it; to answer a repeated request differently, use an `on_tx` lambda with a static counter and
+`inject_to_rx_buffer` (see `uart_mock_ld2412_requery_after_giveup.yaml`).
+
+## YAML build tests
+
+```
+tests/
+├── test_build_components/
+│   └── common/          # Shared bus packages (uart, i2c, spi, etc.)
+│       ├── uart/        # UART at default baud rate
+│       ├── uart_115200/ # UART at 115200 baud
+│       ├── i2c/         # I2C bus
+│       └── spi/         # SPI bus
+└── components/[component]/
+    ├── common.yaml          # Component-only config (no bus definitions)
+    ├── test.esp32-idf.yaml      # config + compile
+    ├── test.esp8266-ard.yaml    # config + compile
+    ├── test-variant.esp32-idf.yaml  # variant test, config + compile
+    ├── validate.esp32-idf.yaml      # config-only (never compiled)
+    └── validate-legacy.esp32-idf.yaml  # config-only variant
+```
+Run them with `script/test_build_components`. Use `-c <component>` to test specific components and
+`-t <target>` for specific platforms.
+
+- **Keep build-test variants few:** one `test.<platform>.yaml` per platform the code actually differs on.
+  A variant per board for platform-identical code only adds CI time.
+- **Config-only test files (`validate.*.yaml`):** use this prefix when a YAML file only needs to exercise
+  schema/validation paths and does not need to be compiled. CI runs `validate.*.yaml` files with
+  `esphome config` only and skips them during compile. The grammar mirrors `test.*.yaml`:
+  - `validate.<platform>.yaml`: base config-only test
+  - `validate-<variant>.<platform>.yaml`: config-only variant
+
+  Use this for deprecated-syntax migration tests, schema edge cases, or platform-specific validation
+  branches where building firmware adds no signal. A component may have any mix of `test.*.yaml` and
+  `validate.*.yaml` files. Validate files never participate in bus grouping; each one runs as its own
+  `esphome config` invocation.
+
+  When a PR's only edits to a component are `validate.*.yaml` files (no source changes, no `test.*.yaml`
+  changes, and the component isn't pulled in as a dependency of another changed component), CI skips the
+  compile stage for that component and only runs config validation. This is decided in
+  `script/determine-jobs.py` via `_component_change_is_validate_only` and surfaced as the
+  `validate_only_components` output that the `test-build-components-split` job consumes.
+
+## Grouping with packages
+
+Components that use shared bus packages can be grouped together in CI to reduce build count. **Never define
+buses (uart, i2c, spi, modbus) directly in test YAML files**; always use packages from
+`test_build_components/common/`.
+
+All includes in test files must go through dict-style `packages:` so that batch grouping works; the
+grouping scripts only understand dict-style packages. Never use list-style packages
+(`packages: [- !include ...]`) or top-level merge keys (`<<: !include common.yaml`). Bus packages are keyed
+by the bus name; the component's `common.yaml` is keyed by the component name (e.g.
+`cst328: !include common.yaml`):
+```yaml
+# test.esp32-idf.yaml - everything included via named packages
+packages:
+  uart: !include ../../test_build_components/common/uart_115200/esp32-idf.yaml
+  my_component: !include common.yaml
+```
+```yaml
+# common.yaml - component config only, NO bus definitions
+my_component:
+  id: my_instance
+
+sensor:
+  - platform: my_component
+    name: My Sensor
+```
+Components that define buses directly are flagged as "NEEDS MIGRATION" and cannot be grouped, increasing
+CI build time. `!extend` and `!remove` in a `test.*.yaml` also make the component ungroupable; add a
+standalone entity with its own id instead.
+
+A test that needs a shared component in a particular state (`esp32_ble` security settings, an option that
+PSRAM rejects) cannot be grouped: the merged block wins and the whole batch fails. Add the component to
+`ISOLATED_COMPONENTS` in `script/analyze_component_buses.py`, or move the option to a `test-<variant>` file.
+
+To verify that all components can be tested together without ID conflicts or configuration issues:
+```bash
+./script/test_component_grouping.py -e config --all
+```
+This tests all components in a single build to catch conflicts that might not appear when testing
+components individually. Use `-e config` for fast configuration validation, or `-e compile` for full
+compilation testing.

.agents/skills/config-validation/SKILL.md

@@ -0,0 +1,124 @@
+---
+name: config-validation
+description: Write or change ESPHome configuration validation - CONFIG_SCHEMA, cv.Schema, cv.typed_schema, custom validator functions, config keys and CONF_ constants, renamed or deprecated keys, and the language schema that script/build_language_schema.py extracts from them. Use whenever a change touches a component's schema or a validator in config_validation.py, automation.py or a component's .py files.
+---
+
+# Configuration validation
+
+ESPHome validates YAML with `voluptuous` schemas built from `esphome/config_validation.py` (`cv`). The
+same schemas are also walked by `script/build_language_schema.py`, which produces the JSON language schema
+that editors and the dashboard use for completion and checking. A validation change is only finished when
+both still work.
+
+## Writing schemas
+
+- **Type hints:** validators are typed like any other function: `def validate_x(config: ConfigType) ->
+  ConfigType:`, with `ConfigType` imported from `esphome.types`.
+- **One schema per model:** when a component's options depend on a `model` key, use
+  `cv.typed_schema({...}, key=CONF_MODEL, default_type=...)` with each model's own keys and defaults, not
+  one schema plus a validator that rejects keys and fills defaults afterwards.
+- **Reuse existing validators:** before writing a custom validator, check for an existing one in
+  `config_validation.py` and compose it in `cv.All(...)` rather than duplicating logic across components.
+  For example, rename a config key with `cv.rename_key(CONF_OLD, CONF_NEW, removed_in="2026.6.0")`, and
+  reject mutually-exclusive keys with `cv.has_at_most_one_key(...)` / `cv.has_exactly_one_key(...)`. See
+  how `api` composes `cv.has_exactly_one_key` + `cv.rename_key`.
+- **Common validators:** `cv.int_`, `cv.float_`, `cv.string`, `cv.boolean`, `cv.int_range(min=0,
+  max=100)`, `cv.positive_int`, `cv.percentage`.
+- **Complex validation:** `cv.All(cv.string, cv.Length(min=1, max=50))`, `cv.Any(cv.int_, cv.string)`.
+- **Platform-specific:** `cv.only_on(["esp32", "esp8266"])`, `esp32.only_on_variant(...)`,
+  `cv.only_on_esp32`, `cv.only_on_esp8266`, `cv.only_on_rp2040`.
+- **Framework-specific:** `cv.only_with_framework(...)`, `cv.only_with_arduino`. On the ESP32, Arduino is
+  built as a component of ESP-IDF, so an ESP-IDF API, sdkconfig option or version floor is there on both
+  frameworks: gate on `esp32.require_idf_version(cv.Version(5, 5, 3), "psram_dma")`, which checks the
+  platform and the ESP-IDF under either framework, as `spi`, `usb_host` and `wifi` do. Reserve
+  `cv.only_with_framework("esp-idf")` for a build that has been tried on Arduino and fails there, and write
+  the actual failure next to it (`sendspin`: arduino-esp32 below ESP-IDF 6.0 bundles a second libsodium,
+  and from 6.0 sendspin-cpp's enumerators collide with `Arduino.h` macros such as `DISPLAY`).
+- **Schema extensions:**
+  ```python
+  CONFIG_SCHEMA = cv.Schema({ ... })
+   .extend(cv.COMPONENT_SCHEMA)
+   .extend(uart.UART_DEVICE_SCHEMA)
+   .extend(i2c.i2c_device_schema(0x48))
+   .extend(spi.spi_device_schema(cs_pin_required=True))
+  ```
+- **Constants:** `esphome/const.py` is frozen; do not add new `CONF_` constants there. Define a
+  component-local constant in the component's own `.py`; for a constant shared by multiple components, add
+  it to `esphome/components/const/__init__.py`. CI (`lint_constants_usage`) fails if the same constant is
+  defined in three or more component files. Constants used in core files (those not under
+  `esphome/components`) may be added to `esphome/const.py` but will require adjustment to the CI check.
+- **Renamed or removed keys:** see the `breaking-changes` skill for `cv.rename_key` and manual
+  deprecation warnings.
+- **Config-only tests:** schema edge cases and deprecated-syntax migrations go in `validate.*.yaml` test
+  files (see the `component-tests` skill).
+
+## Schema extraction
+
+`script/build_language_schema.py` sets `esphome.schema_extractors.EnableSchemaExtraction = True` before it
+imports any component, then walks every `CONFIG_SCHEMA` (and platform schema) recursively. It understands
+`cv.Schema`, `cv.All`, `cv.Any`, dicts of `cv.Required` / `cv.Optional` keys, `.extend()`, `cv.ensure_list`,
+`cv.typed_schema`, registries and the known `cv` validators. **A plain function it cannot see into**: a
+custom validator that builds or picks a schema at validation time hides every key behind it, and the
+dumper silently emits nothing for them. CI only runs the script with `--check`, which does not compare the
+output, so a lost schema is not caught.
+
+When a validator function wraps, chooses or builds a schema (rather than checking a single value), make it
+visible to the dumper:
+
+1. Decorate it with `@schema_extractor("<type>")` from `esphome.schema_extractors`.
+2. At the top of the function, return the data for that type when it is called with `SCHEMA_EXTRACT`
+   instead of a config value. Compare with `is` (`value is SCHEMA_EXTRACT`); older code uses `==`.
+
+```python
+from esphome.schema_extractors import SCHEMA_EXTRACT, schema_extractor
+
+
+@schema_extractor("schema")
+def _outgoing_connection_schema(config: ConfigType | None) -> ConfigType:
+    if config is SCHEMA_EXTRACT:
+        return _OUTGOING_CONNECTION_SCHEMA
+    if config is None:
+        config = {}
+    return _OUTGOING_CONNECTION_SCHEMA(config)
+```
+
+The type names the dumper accepts, and what `SCHEMA_EXTRACT` must return for each, are in the
+`hidden_schemas` branch of `convert()` in `script/build_language_schema.py`. The common ones:
+
+| Type | Return for `SCHEMA_EXTRACT` | Example |
+|------|-----------------------------|---------|
+| `"schema"` | the schema (or a representative one) to walk | `api` `_outgoing_connection_schema`, `restore_state_schema` in `esphome/core/entity_helpers.py`, `model_schema_extractor` in `esphome/components/mipi/__init__.py` |
+| `"enum"` | the dict of options | `sensor_entity_category` in `esphome/components/sensor/__init__.py` |
+| `"one_of"` | the list of options | `cv.one_of` |
+| `"maybe"` | `(validator, key)` for a scalar shorthand | `cv.maybe_simple_value`, `maybe_conf` in `esphome/automation.py` |
+| `"templatable"` | the inner validator | `cv.templatable` |
+| `"automation"` | the automation schema | `automation.validate_automation` |
+
+Points to watch:
+
+- The dumper matches the function object by `repr`, so the decorated function must be the exact object
+  placed in the schema. In a factory that returns a new closure per call, decorate the inner function (as
+  `restore_state_schema` and `cv.maybe_simple_value` do), not the factory.
+- The decorator only records anything when extraction is enabled, and that flag is set before imports, so
+  the decorator must run at import time (module level or inside a factory called at module level).
+- A schema chosen per config value (for example by `model`) has no single answer: return a representative
+  schema, as `model_schema_extractor` does.
+- Plain value checks (`cv.All(cv.string, my_check)`) need nothing: the dumper still sees the surrounding
+  types and keys.
+- `cv.Schema`, `cv.typed_schema`, `cv.ensure_list` and the `cv` helpers are already handled; prefer
+  composing them over a custom function, which also avoids the extractor work.
+
+### Checking the output
+
+Build the language schema into a scratch directory and look at the component's JSON (it takes about ten
+seconds):
+
+```bash
+python3 script/build_language_schema.py --output-path /tmp/schema
+python3 -c "import json; print(json.dumps(json.load(open('/tmp/schema/my_component.json')), indent=1))" | less
+```
+
+Every key a user can write should appear under `config_vars` with a sensible `type`; compare against the
+same output from `dev` when changing an existing schema. If the script prints `Unknown extracted schema
+type` or raises, the decorator type or the value returned for `SCHEMA_EXTRACT` is wrong. Unit tests for
+the dumper live in `tests/script/test_build_language_schema.py`.

.agents/skills/embedded-memory/SKILL.md

@@ -0,0 +1,169 @@
+---
+name: embedded-memory
+description: Choose ESPHome C++ containers and storage that avoid heap use and fragmentation on microcontrollers - std::array, StaticVector with cg.slot_counter, FixedVector, std::unique_ptr byte buffers, StringRef, RAMAllocator, and what to avoid (std::vector growth, std::map, std::deque, std::string, new (std::nothrow)). Use when adding or changing C++ members, containers, buffers, strings or registration lists, or when reducing RAM or flash use.
+---
+
+# Memory on embedded targets
+
+ESP devices run for months with small heaps shared between Wi-Fi, BLE, LWIP, and application code. Over
+time, repeated allocations of different sizes fragment the heap. Failures happen when the largest
+contiguous block shrinks, even if total free heap is still large. We have seen field crashes caused by
+this.
+
+**Heap allocation after `setup()` should be avoided unless absolutely unavoidable.** Every
+allocation/deallocation cycle contributes to fragmentation. ESPHome treats runtime heap allocation as a
+long-term reliability bug, not a performance issue. Helpers that hide allocation (`std::string`,
+`std::to_string`, string-returning helpers) are being deprecated and replaced with buffer and view based
+APIs.
+
+## Container guidelines
+
+1. **Compile-time-known sizes:** use `std::array` instead of `std::vector` when size is known at compile
+   time.
+   ```cpp
+   // Bad - generates STL realloc code
+   std::vector<int> values;
+
+   // Good - no dynamic allocation
+   std::array<int, MAX_VALUES> values;
+   ```
+   Use `cg.add_define("MAX_VALUES", count)` to set the size from Python configuration.
+
+   **For byte buffers:** avoid `std::vector<uint8_t>` unless the buffer needs to grow. Use
+   `std::unique_ptr<uint8_t[]>` instead. It does **not** provide bounds checking or iterator support like
+   `std::vector<uint8_t>`; use it only when you do not need these features.
+   ```cpp
+   // Bad - STL overhead for simple byte buffer
+   std::vector<uint8_t> buffer;
+   buffer.resize(256);
+
+   // Good - minimal overhead, single allocation
+   std::unique_ptr<uint8_t[]> buffer = std::make_unique<uint8_t[]>(256);
+   // Or if size is constant:
+   std::array<uint8_t, 256> buffer;
+   ```
+
+2. **Compile-time-known fixed sizes with vector-like API:** use `StaticVector` from
+   `esphome/core/helpers.h` for compile-time fixed size with a `push_back()` interface (no dynamic
+   allocation).
+   ```cpp
+   // Bad - generates STL realloc code (_M_realloc_insert)
+   std::vector<ServiceRecord> services;
+   services.reserve(5);  // Still includes reallocation machinery
+
+   // Good - compile-time fixed size, no dynamic allocation
+   StaticVector<ServiceRecord, MAX_SERVICES> services;
+   services.push_back(record1);
+   ```
+   Use `cg.add_define("MAX_SERVICES", count)` to set the size from Python configuration.
+
+   **Listener and child-entity registration lists are the most common case, and the most commonly
+   missed.** A `register_*()` method called once per child at code generation time has a count that is
+   known at compile time, so it should never be a `std::vector`. Use `cg.slot_counter()`: it returns a
+   function that each consumer calls once per slot it will occupy, and after every `to_code` has run it
+   emits the define with the final count. When nothing registers, no define is emitted and the storage plus
+   its registration method compile out entirely.
+   ```python
+   # hub component's __init__.py
+   _request_listener_slot = cg.slot_counter("MY_COMPONENT_LISTENER_COUNT")
+
+
+   async def register_listener(hub: MockObj, var: MockObj) -> None:
+       _request_listener_slot()
+       cg.add(hub.register_listener(var))
+   ```
+   When several instances each own a list declared at the same size (one per hub of a `MULTI_CONF`
+   component), pass the owning object as the key, `_request_listener_slot(str(hub))`; the define is then
+   the largest count any one key requested instead of the total.
+   ```cpp
+   #ifdef MY_COMPONENT_LISTENER_COUNT
+     void register_listener(MyComponentListener *listener);
+   #endif
+    protected:
+   #ifdef MY_COMPONENT_LISTENER_COUNT
+     StaticVector<MyComponentListener *, MY_COMPONENT_LISTENER_COUNT> listeners_;
+   #endif
+   ```
+   Request slots from `to_code`, not from a job that runs after `CoroPriority.FINAL`; a late request
+   raises rather than silently undercounting.
+
+3. **Runtime-known sizes:** use `FixedVector` from `esphome/core/helpers.h` when the size is only known at
+   runtime initialization.
+   ```cpp
+   // Bad - generates STL realloc code (_M_realloc_insert)
+   std::vector<TxtRecord> txt_records;
+   txt_records.reserve(5);  // Still includes reallocation machinery
+
+   // Good - runtime size, single allocation, no reallocation machinery
+   FixedVector<TxtRecord> txt_records;
+   txt_records.init(record_count);  // Initialize with exact size at runtime
+   ```
+   This eliminates `_M_realloc_insert` and `_M_default_append` template instantiations (200-500 bytes per
+   instance), needs no upper bound, and is compatible with protobuf code generation when using
+   `[(fixed_vector) = true]`.
+
+4. **Small datasets (1-16 elements):** use `std::vector` or `std::array` with simple structs instead of
+   `std::map`/`std::set`/`std::unordered_map`.
+   ```cpp
+   // Bad - 2KB+ overhead for red-black tree/hash table
+   std::map<std::string, int> small_lookup;
+   std::unordered_map<int, std::string> tiny_map;
+
+   // Good - simple struct with linear search (std::vector is fine)
+   struct LookupEntry {
+     const char *key;
+     int value;
+   };
+   std::vector<LookupEntry> small_lookup = {
+     {"key1", 10},
+     {"key2", 20},
+     {"key3", 30},
+   };
+   // Or std::array if size is compile-time constant:
+   // std::array<LookupEntry, 3> small_lookup = {{ ... }};
+   ```
+   Linear search on small datasets is often faster than hashing or tree overhead, but this depends on
+   lookup frequency; in hot paths the O(1) vs O(n) difference may still matter. It is the heavy containers
+   (`map`, `set`, `unordered_map`) that should be avoided for small datasets unless profiling shows
+   otherwise.
+
+5. **Strings set once from configuration:** use `StringRef` (`esphome/core/string_ref.h`) rather than
+   `std::string`. Code generation passes a string literal that lives in flash for the life of the program,
+   so storing a `std::string` copies it onto the heap for nothing. `StringRef` is a non-owning pointer plus
+   length; it does not copy, and it must only ever refer to storage that outlives it (a string literal, or
+   a buffer owned elsewhere).
+   ```cpp
+   // Bad - heap copy of a literal that is already in flash
+   void set_keys(std::string keys) { this->keys_ = std::move(keys); }
+   std::string keys_;
+   ```
+   ```cpp
+   // Good - no allocation
+   void set_keys(const char *keys) { this->keys_ = StringRef(keys); }
+   StringRef keys_;
+   ```
+
+6. **Avoid `std::deque`:** it allocates in 512-byte blocks regardless of element size, guaranteeing at
+   least 512 bytes of RAM usage immediately. This is a major source of crashes on memory-constrained
+   devices.
+
+7. **Never use `new (std::nothrow)`:** on ESP-IDF exceptions are disabled, so a failed nothrow allocation
+   aborts instead of returning `nullptr`. Use `RAMAllocator` from `esphome/core/helpers.h`; CI rejects
+   `std::nothrow`.
+
+8. **Detection:** look for these patterns in compiler output:
+   - Large code sections with STL symbols (vector, map, set)
+   - `alloc`, `realloc`, `dealloc` in symbol names
+   - `_M_realloc_insert`, `_M_default_append` (vector reallocation)
+   - Red-black tree code (`rb_tree`, `_Rb_tree`)
+   - Hash table infrastructure (`unordered_map`, `hash`)
+
+Callback storage (`CallbackManager` vs `LazyCallbackManager`) and templated registration methods are
+covered in the `automations` skill.
+
+## Where to spend effort
+
+Prioritize container optimization for core components (API, network, logger), widely used components
+(mdns, wifi, ble), and components causing flash size complaints. Avoiding heap allocation after `setup()`
+is always required regardless of component type; the priority list is about effort spent on container
+migration (e.g. `std::vector` to `StaticVector`).

.agents/skills/hardware-test/SKILL.md

@@ -111,11 +111,11 @@ ESP-IDF OTA image: `.esphome/build/<name>/build/firmware.ota.bin`. LibreTiny OTA
   or exits 1 when a build, flash or port step failed.
 
 Per-topic scripts, each described in the matching `reference/recipes/<component>.md`: `online_image/`
-(server, image generator, API driver), `zwave_proxy/` (checksum verifier, device info), `usb_cdc_acm/`
+(server with prefixed misbehaviour cases, PNG, JPEG, BMP and QOI generators, API drivers, first pass timer), `zwave_proxy/` (checksum verifier, device info), `usb_cdc_acm/`
 (host driver, Zephyr driver probe), `rtsp_server/` (protocol client, raw rate reader, measurement set,
 API image grab), `cse7766/` (API observer with ping), `zephyr_coredump/` (reader flow, decode with the
 Zephyr tools), `libretiny_prefs/` (rtc preference harness header), `libretiny_crash/` (BK72xx crash probe header and
-capture decoder), `api_outgoing/` (wake cycle and Wi-Fi drop timing readers), `ld2450/` (zone number driver, loopback harness driver with number and binary sensor states and TX frame summarizer), `api/` (raw plaintext native API client for user defined actions), `dfrobot_sen0395/` (API driver and case runner), `homeassistant_light/` (direct bulb driver, held-report case driver), `usb_audio/` (URL player), `speaker_media_player/` (API player driver, pipeline task case runner), `web_server/` (SSE client, request mix, SSE and HTTPD case runners), `mqtt/` (broker config with TLS, client-certificate and WebSocket listeners, test CA with server and client certificates, transport case runner, flash-and-drive runner, CLI TLS check without a board), `udp/` (datagram recorder, `udp.write` payload checker and case driver), `entity_icons/` (icon and device class readback over the API, MQTT discovery, web_server and dump_config, with a comparer).
+capture decoder), `api_outgoing/` (wake cycle and Wi-Fi drop timing readers), `ld2450/` (zone number driver, loopback harness driver with number and binary sensor states and TX frame summarizer), `api/` (raw plaintext native API client for user defined actions, store_yaml recovery client over Noise or plaintext), `http_request/` (update manifest server, update state driver, request test server, action case driver, concurrency, reboot, update check and OTA drivers, `ota_auth/` OTA request cases with basic auth), `modem/` (button and sensor driver over the plaintext API, also used for the BLE MIDI boards), `dfrobot_sen0395/` (API driver and case runner), `homeassistant_light/` (direct bulb driver, held-report case driver), `usb_audio/` (URL player), `speaker_media_player/` (API player driver, pipeline task case runner), `web_server/` (SSE client, request mix, SSE and HTTPD case runners), `mqtt/` (broker config with TLS, client-certificate and WebSocket listeners, test CA with server and client certificates, transport case runner, flash-and-drive runner, CLI TLS check without a board, MAC suffix runner with per-variant configs for two boards, `shutdown/` shutdown message runner and matcher), `udp/` (datagram recorder, `udp.write` payload checker and case driver), `entity_icons/` (icon and device class readback over the API, MQTT discovery, web_server and dump_config, with a comparer), `uart_sim/` (API driver and flash, capture and drive runner for the simulated UART sensors), `remote_base/` (loopback runner and API log decoder per sent frame).
 
 ## Templates in `$SKILL/templates`
 
@@ -132,7 +132,21 @@ capture decoder), `api_outgoing/` (wake cycle and Wi-Fi drop timing readers), `l
   `globals-storage-dut.yaml`, `usb-audio-waveshare-s3-eth-dut.yaml`, `usb-audio-sendspin-s3-eth-dut.yaml`,
   `cast-pro-1-sendspin-control.yaml`, `aux-2-preprod-stock-dut.yaml`, `aux-2-preprod-usb-audio-dut.yaml`,
   `dfrobot-sen0395-dut.yaml`, `aux-2-preprod-speaker-player-dut.yaml`, `web-server-sse-dut.yaml`,
-  `mqtt-transport-dut.yaml`, `mqtt-tls-dut.yaml`, `udp-write-dut.yaml`, `entity-icon-dut.yaml`.
+  `mqtt-transport-dut.yaml`, `mqtt-tls-dut.yaml`, `udp-write-dut.yaml`, `entity-icon-dut.yaml`,
+  `online-image-s3-eth-dut.yaml`,
+  and the directories `store_yaml/` (package with secrets one directory up, per-board files),
+  `http_request_update/` (manifest URL setter shapes, per-board files), `modem/` (modem only and modem plus
+  Wi-Fi boot checks), `ble_midi/` (fake BLE MIDI instrument and the client), `http_request_action/` (request
+  action package with per-board files and an OTA flash variant) `uart_sim/` (simulator header, common package, per-build parts and run files) and `remote_base/` (common
+  package, transmit script, per-board templates and the ESP8266 software sender).
+
+## When the PR keeps moving
+
+Record every head in the stack (`gh pr view N --json headRefOid`), build, and recheck before flashing; if anything
+moved, wait five minutes and check again so the test runs on a settled head. When a head moved only by merging dev,
+compare the PR's own diff (`git diff $(git merge-base OLD origin/dev) OLD` against the same for NEW); if it is
+identical and the dev commits do not touch the code under test, earlier results still describe it, but rebuild so the
+comment quotes the exact commits. Fetch `origin dev` before computing a merge base.
 
 ## When something looks wrong
 

.agents/skills/hardware-test/reference/boards.md

@@ -50,6 +50,9 @@ Ports are not listed on purpose: they follow the USB hub port and change on ever
   USB replug fixes it. Fallback: `esphome logs --device <name>.local` (API log stream, which drops lines
   while Wi-Fi is off).
 - The ESP32-P4 before silicon revision 3 is bricked by rev3-only firmware (boot loop at bootloader entry).
+- A board left in the ROM bootloader after an upload (BOOT held, or the USB JTAG reset landing in download mode
+  again) starts its app without a replug through `esptool --port <port> --after watchdog-reset chip-id`; flash a
+  minimal Wi-Fi plus API image first when recovering from a crash-looping one.
 
 ## RP2040 / RP2350 (Pico W, Pico 2 W)
 
@@ -83,6 +86,8 @@ Ports are not listed on purpose: they follow the USB hub port and change on ever
   from `~/.platformio/packages/toolchain-gccarmnoneeabi/bin/` (`raw_firmware.elf`, not `firmware.elf`).
   Fault dumps print PC, LR and R0 to R3.
 - The AmebaZ2 SDK sets `LWIP_TCPIP_CORE_LOCKING 0`.
+- The `rtw_is_adapter_up` bus fault after `Disabling AP` happens on LibreTiny 1.13.0 and 1.14.0 alike
+  (2 of 4 and 3 of 3 runs in #21531); run at least four rounds per build before blaming a PR for it.
 
 ## BK7238, WB3S, CB3S (BK7231N) (LibreTiny bk72xx, CH340, RTS on CEN)
 
@@ -113,6 +118,13 @@ Ports are not listed on purpose: they follow the USB hub port and change on ever
   SDK's saved register banks are usable. Code is compiled Thumb, the SDK assembly is ARM; in an exception
   handler only r0 to r3 are shared by all modes. BK7231N ITCM is `0x3FEE00..0x400000` (4.5 KB); a `.noinit`
   placed there after `.itcm.code` works for a crash record.
+- Pick the board that matches the module's flash layout. A NiceMCU XH-WB3S (BK7238) needs `board: xh-wb3s` (plain
+  Beken layout, RF calibration at 0x1E0000). Built as `generic-bk7238-tuya` (calibration at 0x1E3000, download
+  partition over 0x1E0000) the firmware boots and may even join a known network, but scans come back empty and its
+  AP is invisible to clients, with no warning in the log. If scans are empty, check the layout first.
+- The BK7238 bootloader prints `App verify failed! Need to recovery factory firmware.` on every boot with any
+  build; it is harmless. Even with the right layout an ESP32 client did not see the BK7238 fallback AP on
+  LibreTiny 1.13.0 or 1.14.0, so the wifisave hammer cannot run on it.
 - Callbacks run on the `async_tcp` task and log lines can lose a byte. Wi-Fi connect takes tens of ms;
   radio enable 160 to 278 ms, disable about 200 ms.
 

.agents/skills/hardware-test/reference/recipes/README.md

@@ -4,17 +4,20 @@ Each flow names the pull requests it was used on, so their result tables show wh
 
 One file per component, named after it:
 
-- [`api.md`](api.md): noise handshake and OTA encryption on the ESP8266, outgoing connection wake timing on deep sleep and after a Wi-Fi drop, user defined action listing and execution with a raw plaintext client
+- [`api.md`](api.md): noise handshake and OTA encryption on the ESP8266, outgoing connection wake timing on deep sleep and after a Wi-Fi drop, user defined action listing and execution with a raw plaintext client, store_yaml recovery over Noise or plaintext
+- [`ble_midi.md`](ble_midi.md): BLE MIDI client against a fake ESP32 instrument, every packet shape and the send actions
 - [`captive_portal.md`](captive_portal.md): fallback AP and /wifisave hammer
 - [`cse7766.md`](cse7766.md): Sonoff S31 parser check over the API
 - [`espnow.md`](espnow.md): ESP8266 broadcast receive
 - [`hdmi_cec.md`](hdmi_cec.md): two-board wire
+- [`http_request.md`](http_request.md): update manifest URL changed at run time, every setter shape, a swap under a running check, non-blocking request actions with loop gap, queue, cancel and reboot checks, OTA request cases with basic auth and MD5 sources
 - [`ld2450.md`](ld2450.md): polygon zones, zone number round trip through the radar's readback, ESP8266 loopback harness without a radar
 - [`ld2412.md`](ld2412.md): output pin level through the basic configuration command, held back writes
 - [`libretiny.md`](libretiny.md): rtc preference emulation, BK72xx crash report on the next boot
-- [`mqtt.md`](mqtt.md): broker bench, QoS, reconnect, crash log over MQTT, WebSocket transports with a local and a public broker, TLS trust sources (own CA, client certificate, certificate bundle, no verification) with the CLI check
+- [`modem.md`](modem.md): PPPoS modem component boot check with no modem attached, loop blocking and disable timing
+- [`mqtt.md`](mqtt.md): broker bench, QoS, reconnect, crash log over MQTT, WebSocket transports with a local and a public broker, TLS trust sources (own CA, client certificate, certificate bundle, no verification) with the CLI check, MAC suffix topics and will messages on two boards, shutdown message delivery on restart, disable and deep sleep
 - [`nrf52.md`](nrf52.md): dongle board support
-- [`online_image.md`](online_image.md): stalled and malformed downloads
+- [`online_image.md`](online_image.md): stalled and malformed downloads, PNG, JPEG, BMP and QOI decode time per loop pass with large images and small buffers
 - [`openthread.md`](openthread.md): border router on the S3 host plus H2 RCP
 - [`ota.md`](ota.md): overlap and lock, timeouts, deflate
 - [`rtsp_server.md`](rtsp_server.md): camera streaming on the Waveshare ESP32-S3-ETH
@@ -31,3 +34,7 @@ One file per component, named after it:
 - [`dfrobot_sen0395.md`](dfrobot_sen0395.md): SEN0395 mmWave command flow, module quirks and overlong lines through its error reply
 - [`speaker.md`](speaker.md): media player pipeline tasks on the Connect AUX-2 with the `speaker` platform player
 - [`web_server.md`](web_server.md): state SSE stress and the HTTPD task stack on web_server_idf, entity icon and device class readback
+- [`uart_sim.md`](uart_sim.md): UART sensors simulated on a classic ESP32 with loopback and cross-wired UARTs
+- [`gpio_expander.md`](gpio_expander.md): interrupt driven expander inputs, loop idle checks, testing a device out of reach
+- [`led_strip.md`](led_strip.md): forcing a failed strip allocation, RMT symbols on the S3
+- [`remote_base.md`](remote_base.md): remote_receiver protocol decoding through a one-pin loopback on ESP32 and ESP8266, no IR hardware

.agents/skills/hardware-test/reference/recipes/api.md

@@ -40,3 +40,19 @@ length delimited strings and `int[]` as repeated sint32 (field 7); `ALLOW_MISSIN
 defaults) and `templates/api-actions-strings-dut.yaml` (required strings) log what arrived. String arguments are
 `StringRef`s that are not null terminated: copy them with `snprintf("%.*s", size, c_str())` into a buffer before
 logging, and do not pass `%.*s` to `logger.log`, whose validator counts it as one pattern.
+
+## store_yaml recovery over the native API (#16445)
+
+`scripts/api/syrecover.py HOST KEY OUTDIR [ORIGDIR]` (KEY `-` for plaintext) builds the client library's
+`APINoiseFrameHelper` around a stub `APIConnection` subclass (the compiled helper type checks the connection, and
+`ConnectionParams` needs every field), sends Hello, Authentication (never answered without a password) and
+`DeviceCapabilitiesRequest`, then `GetYamlRequest` (161), collects `GetYamlResponse` (162) frames by field number
+(1 data, 2 done, 3 total_size, 4 encoding) until done, requests again, decompresses with `compression.zstd` (Python 3.14, or the `zstandard` package) and
+unpacks the envelope with the tree's `unpack_envelope` (run with `PYTHONPATH` on that tree). Configs in
+`templates/store_yaml/`: entry files in `boards/`, a package one directory above them with `!secret` and an inline
+OTA password,
+so the redaction and the `..` paths are exercised. Repetitive YAML compresses into one chunk even at 22 KB; a 6 KB
+include of random base64 lines with `include_secrets: true` gives four chunks (1390 byte packets). Rough edge at
+0bd3738430: `include_secrets: true` fails the build when a package in another directory uses `!secret` and has no
+`secrets.yaml` beside it; a copy next to the package works around it.
+Results: https://github.com/esphome/esphome/pull/16445#issuecomment-6335435397

.agents/skills/hardware-test/reference/recipes/ble_midi.md

@@ -0,0 +1,19 @@
+# ble_midi
+
+## BLE MIDI client against a fake instrument (#18508)
+
+No MIDI hardware needed: `templates/ble_midi/midiperi.yaml` turns a second ESP32 (an S3 with ethernet, so the
+API stays reachable while BLE is busy) into a BLE MIDI peripheral with `esp32_ble_server`: the MIDI service
+advertised, the MIDI characteristic with read, write, write without response and notify, an `on_write` that logs
+the bytes, buttons that notify hand built packets (channel messages, two messages in one packet, running status,
+SysEx in one packet, split over two, oversize, cut short, clock, aftertouch, orphan data bytes, a 50 packet burst)
+and a restart button for the disconnect test. `midicli.yaml` is the component test config on a classic devkit with
+`wifi:`, `api:`, the peripheral's BLE address in the `peri_mac` substitution, given on the command line with
+`-s peri_mac <mac>` (the base MAC plus 2, printed by `esp32_ble` in `dump_config`), `on_sysex` and `on_message` lambdas that print the bytes, debug sensors and out of range send
+buttons. Drive both over the plaintext API with `scripts/modem/mdrive.py HOST LOGFILE press:NAME wait:S` 3 s
+apart and compare the peripheral's `TX` and `RX` lines with the client's decoded lines. Gotcha:
+`ble_server.characteristic.set_value` latches the value of its first play and re-applies it before every notify,
+so the peripheral script sets the value from a lambda (`id(midi_io).set_value(std::vector<uint8_t>(pkt));
+id(midi_io).notify();`). At VERBOSE the client logs four lines per message, so a burst shows an
+`esp32_ble took a long time` of a couple of seconds that is the serial port, not the parser.
+Results: https://github.com/esphome/esphome/pull/18508#issuecomment-6348272717

.agents/skills/hardware-test/reference/recipes/gpio_expander.md

@@ -0,0 +1,13 @@
+# GPIO expanders
+
+## Interrupt driven base class (#21469, #21468)
+
+Any device with a PCF8574 input expander whose INT line goes to an ESP32 GPIO works; the inputs should change
+during the test (end switches on a gate or cover are ideal). Before flashing, check with `nm` that
+`gpio_expander::interrupt_pin_isr` lands in IRAM (0x4008xxxx on a classic ESP32). Then confirm `dump_config` lists
+the interrupt pin and the input states match reality, and that the `pcf8574` entry is absent from the runtime stats
+while idle (its loop is disabled). Drive the inputs (for a cover, `cover_command(position=1.0)` and then `0.0` over
+the API, waiting for each end state): every input edge should log, `pcf8574` should show a small count in that
+minute's stats and vanish in the next. For a device that cannot be reached physically, boot the same code on a
+bench board first and keep a merge base image built as the fallback; an ESP32 bootloader without rollback support
+leaves only `safe_mode` to recover a bad image.

.agents/skills/hardware-test/reference/recipes/http_request.md

@@ -0,0 +1,48 @@
+# http_request
+
+## Update manifest URL changed at run time (#21373)
+
+`scripts/http_request/manifest_server.py PORT [DELAY]` serves `stable.json`, `beta.json` and `slow.json` (answered
+after DELAY s) with both chip families and logs every request. `templates/http_request_update/upd-common.yaml`
+(secret `manifest_base` is the server URL) has the update entity on the stable manifest, buttons for every
+`set_source_url` shape (`std::string` local, literal, temporary, `.c_str()` of a local, the current URL back in),
+`Check Then Swap` (ESP32 task path: check the slow manifest and replace the URL in the same lambda), `Arm Swap In
+Check` plus a `logger: on_message` lambda that replaces the URL from inside the inline check (ESP8266; the log line
+from inside `on_message` is swallowed, judge by the next check's result), and a `Heap` button. `upd-base-common.yaml`
+keeps only the shapes that compile on a tree with the bare pointer setter (for the base build, copy a board file
+and point its `common` include at `../upd-base-common.yaml` with a different `name:`); `boards/upd8266.yaml` and `boards/upd32.yaml` are the
+per-board files, one directory below the package their `../` include points at. Drive with `scripts/http_request/updrive.py HOST KEY LOGFILE press:NAME wait:S`, which prints every
+update state (latest version, summary). The ESP8266 Arduino client gets `Error reading manifest: -1` on roughly a
+quarter of the fetches from this Python server on every build, so count checks against the base before blaming a
+PR. `http_request` on the ESP8266 needs `verify_ssl: false` even for plain http URLs.
+Results: https://github.com/esphome/esphome/pull/21373#issuecomment-6337544933
+
+## Non-blocking request actions (#21486)
+
+`scripts/http_request/hsrv.py PORT` serves `/fast`, `/slow?d=S`, `/big` and `/chunked` (`n` bytes at `rate` B/s,
+deterministic bytes `(i*7+3)&0xFF`), `/post` (echoes length and byte sum), `/redirect` and `/hang`.
+`templates/http_request_action/hr-common.yaml` (substitutions `name`, `bufsize`; per-board files `hr32.yaml` and
+`hr66.yaml`, `-s name hr32base` on the command line for the base build; `hrota32.yaml` adds the `otaflash` action
+for `otadrive.py`) has API actions `req_get`, `req_post`, `burst` (N runs of a `mode: parallel` script on one action),
+`cancel` (`mode: single` script run twice, then `script.stop`) and `reboot_during`, and logs
+`RESULT n status len sum took maxgap`, where maxgap is the longest main loop gap during the request (an
+`interval: 1ms` lambda plus the time since its last pass, so a fully blocking request counts too). Drive with
+`scripts/http_request/hrdrive.py IP http://HOST:PORT LABEL [https]` (`ONLY=case,case` and `ROUNDS=n` env; the
+`cross` case runs two actions on one component). `burstonly.py IP URL N` checks concurrency against the server's
+completion times, `reboottest.py IP URL` times `App.safe_reboot()` with requests in flight, `updcheck.py IP LABEL N`
+runs N update checks over the API (for real device configs that use `update: platform: http_request`) and
+`otadrive.py` triggers `ota.http_request.flash` (the API variable is `sum`; `md5` is a reserved variable name).
+`execute_service` must be awaited on current aioesphomeapi or nothing is sent. Keep the ESP32 `bufsize` at 32 KB or
+less: the action copies the body into a `std::string` while the buffer is still held, and 80 KB aborts with out of
+memory on dev and on the PR alike. `http_request` refuses BK72xx and RTL87xx at validation.
+
+## OTA request cases with basic auth (#21556)
+
+`scripts/http_request/ota_auth/`: `ho-common.yaml` (API actions for `ota.http_request.flash` with `md5_url`, `md5`
+and basic auth, `flash()` with no URL, an invalid URL, the `md5_computed()`/`md5_expected()` getters), `fwsrv.py`
+(serves a www dir, `/auth/` needs `otauser:otapass`), `hodrive.py` (11 cases, reconnects after each good flash) and
+`run.sh BUILD BOARD` (compiles from `TREE`, serial flash, serves that exact image; `TREE`, `P32` or `P66`, `SC` and `HOST_IP` in
+the environment). On the D1 mini the authenticated MD5 read fails
+with `Error reading MD5: -1` on dev too, and stored credentials carry over to the next flash on both builds.
+
+Results: https://github.com/esphome/esphome/pull/21556#issuecomment-6381067561

.agents/skills/hardware-test/reference/recipes/led_strip.md

@@ -0,0 +1,10 @@
+# Addressable LED strips (#21447, #21445)
+
+- To exercise a failed strip allocation, use a classic ESP32 without PSRAM with `num_leds: 65535` and
+  `channel_colors: GRBW` (262 kB, bigger than any internal block). `num_leds` is uint16 in the drivers, larger
+  values wrap. On dev such a strip boot-loops the board.
+- Never force allocation failures by hogging RAM instead: Wi-Fi then aborts in `operator new` or the loop task
+  trips the task watchdog.
+- On the ESP32-S3 give each RMT strip `rmt_symbols: 48`, or the first channel takes all 192 symbols.
+- A board with a few WS2812 on a pin already defined as a light can be extended with `id: !extend <light id>` to
+  add effects; restore its own config afterwards.

.agents/skills/hardware-test/reference/recipes/modem.md

@@ -0,0 +1,18 @@
+# modem
+
+## PPPoS modem component with no modem attached (#6721)
+
+A boot check on a classic ESP32 devkit with the modem UART pins left open. `templates/modem/modem-common.yaml` is
+the package (`api:`, debug `loop_time` and `free` sensors, uptime, buttons for `modem.enable`, `modem.disable` and
+`modem.send_at`, a 20 s alive log); `modemo.yaml` is the modem only build with `reboot_timeout: 2min` and the
+component test config's `on_boot` block at `priority: -100` (earlier, `modem.send_at` runs before the modem's
+`setup()` and logs `Modem DCE not ready`); `modemw.yaml` adds `wifi:` with `reboot_timeout: 0s` so the API stays
+reachable. Expect `Modem not responding at 115200 baud` once per pass with each pass blocking the loop about 1 s,
+the reboot at the configured timeout (`SW_CPU_RESET`), and a `modem.disable` with no modem answering to block
+the loop about 37 s (`Failed to leave data mode`, then the radio off timeout) without tripping the 30 s watchdog
+floor because the component feeds it in between. Drive the Wi-Fi build with
+`scripts/modem/mdrive.py HOST LOGFILE list press:NAME wait:S` (plaintext API; prints sensor and text sensor
+states, saves the device log). Before 8cc8df4a1f a modem only config with `api:` failed validation because a
+config dependent `AUTO_LOAD` added `network` after the dependency check; an explicit `network:` block was the
+workaround. Results: https://github.com/esphome/esphome/pull/6721#issuecomment-6347395159 and
+https://github.com/esphome/esphome/pull/6721#issuecomment-6348520416

.agents/skills/hardware-test/reference/recipes/mqtt.md

@@ -51,3 +51,31 @@ context rejects a CA that lacks a `keyUsage` extension (`CA cert does not includ
 on the device accepts it, which is why the test CA carries `basicConstraints` and `keyUsage`. A broker kill shows as one
 `Last esp-tls error: 0x8008` on every build, TLS or not. Results:
 https://github.com/esphome/esphome/pull/14318#issuecomment-6311863827
+
+## MAC suffix topics (#9315)
+
+`scripts/mqtt/mac_suffix/run.sh X BUILD` compiles (from `TREE`) and flashes variant X (a defaults, b custom `topic_prefix`, c explicit topics,
+d no suffix) of BUILD (`pr` or `base`) on an ESP32 and a D1 mini, records the broker with
+`mosquitto_sub -F '@H:@M:@S r=%r %t %p'`, presses Restart through the `cmd_t` found in the discovery payloads, then
+holds both chips in the bootloader for 40 s (`esptool --after no-reset chip-id`) so the broker sends the will; a
+plain reset only gives "session taken over" and no will. The two boards' configs live in separate folders because
+variants a to c share a `name` (the MAC suffix tells them apart), and the same name in one folder shares a build
+directory. A D1 mini CH340 can garble the serial capture after an RTS reset, so read its topics from the broker.
+`esphome logs --device MQTT` on a current tree needs paho-mqtt 2.x; an older venv with 1.6.1 fails, so use a venv
+made from the tree's requirements.txt. A Restart press never delivered the shutdown message on any build tested.
+
+## Shutdown message delivery (#21566)
+
+`scripts/mqtt/shutdown/`: `sd-common.yaml` sets birth/will/shutdown payloads `hello`/`dead`/`bye` so the shutdown
+message and the will cannot be confused, and takes `restart`, `disable` and `sleep` on `<name>/cmd`.
+`sdrun.sh V N` (`TREE`, `P32` and `P66` in the environment) compiles and flashes the ESP32 and D1 mini, restarts each N times, disables MQTT once and deep sleeps the ESP32 three
+times (never the D1 mini: without the GPIO16 to RST wire it never wakes); `match.py OUT` counts a `bye` within 5 s of
+each command. Dev: ESP32 0/10 restarts, D1 mini 1 to 3/10. Put the broker address (`broker_ip`) in the secrets.yaml next to
+the included package, and quote `"${name}/status"` inside flow mappings.
+
+Results: https://github.com/esphome/esphome/pull/21566#issuecomment-6381448061
+
+Retest at c9cf3c2813 (shutdown message sent at QoS 1, teardown waits for the broker's ack): 10/10 restarts on both
+boards and 3/3 deep sleeps, against 1/10, 3/10 and 0/3 on base. An earlier 100 ms hold gave 3/10, 6/10 and 2/3, so
+judge a delivery fix by counts over 10 or more restarts, never one or two presses.
+Results: https://github.com/esphome/esphome/pull/21566#issuecomment-6382076613

.agents/skills/hardware-test/reference/recipes/online_image.md

@@ -18,3 +18,47 @@ ends with `Error downloading image: -28679` (EAGAIN) after the transport timeout
 and cut chunked cases and every later update logs `Image already being updated.`, so a baseline run needs
 a reboot per group, the driver's 45 s cap per case, and an outer `timeout` that covers the whole list.
 A case expected as `any` passes on `finished` or `error`, never on no outcome.
+
+## PNG decode time limit (#21364, #21391)
+
+The same server also serves any plain file in its www directory by name (404 for a missing file or a path
+outside it). `scripts/online_image/make_big_pngs.py <www>` (after `make_images.py`) writes a 320x240 gradient (618 B, also as `ok.png` so the misbehaviour cases carry a
+real multi-row PNG), a flat 640x480 PNG (1.6 KB for 307200 pixels), 231 KB of noise and a JPEG. The device is
+`templates/online-image-dut.yaml` with `type: BINARY` on the main image (a 640x480 RGB565 does not fit a classic
+ESP32 without PSRAM) and `debug` loop time and heap sensors at 2 s; the 256 byte buffer image stays RGB565, so do
+not point big images at it (out of memory on every build). `scripts/online_image/pngdrive.py HOST KEY LOGFILE
+BASEURL LABEL` (default case set `png`; `CASES=short` for the six case list) calls the actions, waits for the
+finished or error line and reports the case time, the highest loop time and the lowest free heap. The per pass `online_image.image took a long time` values come from
+the serial capture and only print on a new maximum, so read the first fetch of each image. The pass that decodes a
+whole buffer is the decoder cost (before the change 510 ms gradient and 1980 ms flat, after it 77 and 140 ms); for
+the noisy PNG the longest pass is the blocking 64 KB read on both builds (0.7 to 3.7 s). Alternate PR, base, PR
+for a timing claim.
+
+For a replacement PR, compare against the parent PR head, not dev. `CASES=noise` fetches the noisy PNG three times
+(run it right after a boot, alternating builds); `CASES=end` runs flat and gradient as plain, `noclen_`, `chunked_`,
+`half_` and `stall_` through the prefixed server cases below.
+
+## JPEG decoded as it downloads (#21342)
+
+`oi_server.py` serves prefixed cases on any file: `half_<file>` (first half, matching Content-Length),
+`noclen_<file>` (close delimited), `chunked_<file>` (1 KB chunks), `stall_<file>` (a third, then the socket held
+40 s). `scripts/online_image/make_jpegs.py <www>` writes noise, gradient, 12 KB EXIF and progressive JPEGs. Device:
+`templates/online-image-dut.yaml` with `type: BINARY` on both entities and `buffer_size: 4096` on the second;
+`CASES=jpeg scripts/online_image/pngdrive.py HOST KEY LOGFILE BASEURL LABEL` runs the JPEG list; every case set
+prints case time, highest loop time and lowest free heap. Before the change a JPEG needs a contiguous block of its file size (233 KB never
+allocates on a classic ESP32, and after a few failures 69 KB does not either until a reboot); after it the 64 KB
+buffer stays 64 KB and 4 KB is enough for a baseline JPEG. The first pass still reads up to `buffer_size` in one
+blocking read, so the 64 KB entity blocks 1.7 to 4 s and the 4 KB one 272 ms at most; judge the decode budget on
+the small buffer entity. Already on dev: a JPEG without Content-Length (plain or chunked) ends with `Could not open
+image for decoding: 4` then `Invalid type`.
+
+## BMP and QOI decode time limit (#21392)
+
+Use a wired board (`templates/online-image-s3-eth-dut.yaml`: Waveshare ESP32-S3-ETH, W5500, PSRAM, `type: RGB565`)
+rather than a classic ESP32 on Wi-Fi: with a weak link the reads block up to the 4.5 s timeout and the per case loop
+numbers are noise. `scripts/online_image/make_bmp_qoi.py <www>` writes 1 bit BMP 640x480, 8 and 24 bit BMP 320x240,
+QOI 640x480 one colour (all runs) and checkerboard, QOI 320x240 gradient and noise. `pngdrive.py` with
+`CASES=bmpqoi` runs each as plain, `noclen_`, `chunked_`, `half_` plus stalls. The `loop_time` sensor sits at 75
+to 160 ms even for a 16x16 image, so for decode passes use `scripts/online_image/firstpass.py HOST KEY BASEURL
+LABEL IMAGE...`: it reboots the device before each fetch and reports that fetch's `took a long time` maximum. A
+`half_` file with matching Content-Length reports `download finished` with "not caching it" on dev too.

.agents/skills/hardware-test/reference/recipes/remote_base.md

@@ -0,0 +1,30 @@
+# remote_base / remote_receiver
+
+## Loopback on one pin, no IR hardware (#21560 to #21563)
+
+`templates/remote_base/`: `rb-common.yaml` (API action `tx_all`, a dumper registered from `on_boot` through
+`register_dumper()` standing in for an external component), `rb-txscript.yaml` (14 `remote_transmitter.transmit_*`
+frames with `tx <proto>` log markers), `rb32.tmpl` and `rb66.tmpl` (`@V@` is the build label).
+`scripts/remote_base/run.sh LABEL B N` (`TREE`, `P32` or `P66`, `SC` in the environment) writes the run file from the
+template, builds and flashes it from `TREE` (`esphome run --no-logs`), finds the IP and runs `rbdrive.py IP N`, which segments the API log by the
+markers and prints what each sent frame decoded as.
+
+- ESP32: `remote_transmitter` and two `remote_receiver`s on the same GPIO (`allow_other_uses`), carrier 100%, no
+  inversion. RMT reads its own output, every frame decodes. Second receiver `dump: [nec, raw]` shows raw runs only
+  where no primary matched.
+- ESP8266: the transmitter sends with interrupts off, so a receiver on the same pin sees nothing. `softtx.hpp`
+  bit-bangs frames built with each protocol's `encode()` with interrupts on, and an `interval` sets the pin back to
+  OUTPUT after the receiver's setup. Bit-banged NEC decodes as NEC about 2 of 3 times (timing), on every build.
+- Dish never decodes in loopback on dev either. Pronto matches every frame, so raw on `dump: all` never prints.
+- API log lines can drop under bursts; rerun with more rounds before calling a shortfall.
+- `register_dumper()` returns nothing, so the config logs `ext dumper register called`, not whether it stuck. On dev
+  the list has a fixed size and a dumper beyond it is dropped with `No dumper slot` (ERROR); judge by that line and by
+  `ext dumper saw` counts in the `rbdrive.py` table, which only appear when the dumper is really registered.
+  `rbdrive.py` exits 1 when the external dumper saw no frame, so a build that rejects the registration fails the
+  drive; set `EXPECT_REGISTER=1` on `run.sh` to also fail the run on any `No dumper slot` line.
+
+Results: https://github.com/esphome/esphome/pull/21563#issuecomment-6381304387
+
+Retest of the moved stack with 5 rounds per build: https://github.com/esphome/esphome/pull/21563#issuecomment-6382075726.
+A D1 mini CH340 sometimes fails the 460800 baud upload; esphome retries at 115200 by itself, but a runner that reads
+only the last line sees "Failed to upload" first, so rerun before blaming the build.

.agents/skills/hardware-test/reference/recipes/uart_sim.md

@@ -0,0 +1,19 @@
+# Simulated UART sensors on a classic ESP32 (#21449, #21446)
+
+No sensor needed. `templates/uart_sim/sim.hpp` (included with `esphome: includes:`) has frame builders for wts01,
+ds1603l, sm300d2 and ufm01, `sim::loopback(id(uart))` (ESP-IDF `uart_set_loop_back` on the component's own UART),
+`sim::send()` that logs the bytes, and `sim::probe`, a SEN0465 responder polled from a 10 ms interval. The run files
+`fra.yaml`, `frb.yaml` and `frc.yaml` set the name and include `fr-common.yaml` plus their `fr*-parts.yaml` as
+packages; copy them with a different name (`frabase` and so on) for the base build.
+
+- Push-style sensors: put the component on a UART, switch it to loopback in `on_boot`, and inject frames from
+  template buttons (`fra-parts.yaml`, `frb-parts.yaml`). ufm01 enters its active stream by itself when a 1 s
+  interval streams frames during its 2 s startup wait.
+- Request/response sensors: cross-wire two UARTs through the GPIO matrix (UART1 tx 25 rx 26, UART2 tx 26 rx 25,
+  `allow_other_uses: true` on all four pins); the second UART runs the simulated device (`frc-parts.yaml`).
+  A select sets the probe's misbehaviour (tail, FF tail, silent, split, wrong gas).
+- Polled readers: run `update()` from a script and log `micros()` around it to measure blocking
+  (sm300d2 on dev blocks about 104 ms on an empty buffer).
+- The logger stays on UART0, so three builds cover five components. Drive with `scripts/uart_sim/frdrive.py HOST
+  LOG press:NAME sel:NAME=OPT num:NAME=V wait:S`, or `scripts/uart_sim/run.sh NAME TREE SECONDS OPS...` (build
+  and flash from TREE with `esphome run --no-logs`, capture, drive; `PORT` and `SC` in the environment).

.agents/skills/hardware-test/scripts/api/syrecover.py

@@ -0,0 +1,221 @@
+"""Pull the store_yaml blob off a device over the native API (plaintext or Noise) and unpack it.
+
+usage: syrecover.py HOST KEY OUTDIR [ORIGDIR]   (KEY of - means a plaintext API)
+Prints the capability flag, chunk count, advertised and received sizes, encoding, the envelope's files, and
+when ORIGDIR is given compares each recovered file with ORIGDIR/<path>.
+Run with PYTHONPATH pointing at the esphome tree that has store_yaml (for unpack_envelope). Needs Python 3.14
+(compression.zstd) or the zstandard package.
+"""
+
+import asyncio
+from pathlib import Path
+import sys
+import time
+
+from aioesphomeapi import api_pb2
+from aioesphomeapi._frame_helper.noise import APINoiseFrameHelper
+from aioesphomeapi._frame_helper.plain_text import APIPlaintextFrameHelper
+from aioesphomeapi.connection import APIConnection, ConnectionParams
+from aioesphomeapi.zeroconf import ZeroconfManager
+
+from esphome.components.store_yaml import unpack_envelope
+
+try:
+    from compression import zstd
+except ImportError:
+    try:
+        import zstandard as zstd
+    except ImportError:
+        sys.exit("needs Python 3.14 (compression.zstd) or the zstandard package")
+
+HOST, KEY, OUT = sys.argv[1], sys.argv[2], Path(sys.argv[3])
+ORIG = Path(sys.argv[4]) if len(sys.argv) > 4 else None
+HELLO, HELLO_RESP, CONNECT, CONNECT_RESP = 1, 2, 3, 4
+DISCONNECT_REQ = 5
+FATAL = -1  # queued by report_fatal_error so a waiting expect() stops at once
+CAPS_REQ, CAPS_RESP = 149, 150
+GET_YAML_REQ, GET_YAML_RESP = 161, 162
+
+
+def decode_varint(buf: bytes, pos: int) -> tuple[int, int]:
+    value = shift = 0
+    while True:
+        b = buf[pos]
+        pos += 1
+        value |= (b & 0x7F) << shift
+        if not b & 0x80:
+            return value, pos
+        shift += 7
+
+
+def fields(payload: bytes) -> dict[int, list]:
+    out: dict[int, list] = {}
+    pos = 0
+    while pos < len(payload):
+        key, pos = decode_varint(payload, pos)
+        fnum, wt = key >> 3, key & 7
+        if wt == 0:
+            v, pos = decode_varint(payload, pos)
+        elif wt == 2:
+            ln, pos = decode_varint(payload, pos)
+            v = payload[pos : pos + ln]
+            pos += ln
+        elif wt == 5:
+            v = payload[pos : pos + 4]
+            pos += 4
+        else:
+            raise ValueError(f"wire type {wt}")
+        out.setdefault(fnum, []).append(v)
+    return out
+
+
+async def main() -> None:
+    loop = asyncio.get_running_loop()
+    queue: asyncio.Queue[tuple[int, bytes]] = asyncio.Queue()
+    errors: list[Exception] = []
+
+    class Shim(APIConnection):
+        """Only the two callbacks the frame helper uses; the real connection logic is bypassed."""
+
+        def process_packet(self, msg_type_proto, data) -> None:
+            queue.put_nowait((msg_type_proto, bytes(data)))
+
+        def report_fatal_error(self, err: Exception) -> None:
+            errors.append(err)
+            queue.put_nowait((FATAL, b""))
+
+    params = ConnectionParams(
+        addresses=[HOST],
+        port=6053,
+        password=None,
+        client_info="syrecover",
+        keepalive=15,
+        zeroconf_manager=ZeroconfManager(),
+        noise_psk=None if KEY == "-" else KEY,
+        expected_name=None,
+        expected_mac=None,
+        timezone=None,
+        provide_time=True,
+    )
+    shim = Shim(params, None, False, HOST)
+    if KEY == "-":
+        helper = APIPlaintextFrameHelper(
+            connection=shim, client_info="syrecover", log_name=HOST
+        )
+    else:
+        helper = APINoiseFrameHelper(
+            connection=shim,
+            noise_psk=KEY,
+            expected_name=None,
+            expected_mac=None,
+            client_info="syrecover",
+            log_name=HOST,
+        )
+    await loop.create_connection(lambda: helper, HOST, 6053)
+    try:
+        await asyncio.wait_for(helper.ready_future, 10)
+    except Exception as exc:  # noqa: BLE001
+        sys.exit(f"handshake failed: {exc!r} {errors}")
+    print(f"transport {'plaintext' if KEY == '-' else 'noise'}: handshake done")
+
+    async def expect(msg_type: int, timeout: float = 10) -> bytes:
+        while True:
+            t, d = await asyncio.wait_for(queue.get(), timeout)
+            if t == msg_type:
+                return d
+            if t == FATAL:
+                sys.exit(
+                    f"transport error while waiting for message {msg_type}: {errors}"
+                )
+            if t == DISCONNECT_REQ:
+                sys.exit(
+                    f"device closed the connection while waiting for message {msg_type}"
+                )
+            print(f"skipping message type {t} while waiting for {msg_type}")
+
+    hello = api_pb2.HelloRequest(
+        client_info="syrecover", api_version_major=1, api_version_minor=10
+    )
+    helper.write_packets([(HELLO, hello.SerializeToString())], False)
+    await expect(HELLO_RESP)
+    # Sent like the client library does; a device without a password never answers it
+    helper.write_packets(
+        [(CONNECT, api_pb2.AuthenticationRequest().SerializeToString())], False
+    )
+    helper.write_packets(
+        [(CAPS_REQ, api_pb2.DeviceCapabilitiesRequest().SerializeToString())], False
+    )
+    caps = fields(await expect(CAPS_RESP))
+    supported = None
+    if 7 in caps:
+        supported = fields(caps[7][0]).get(1) == [1]
+    print(f"capabilities: store_yaml field present={7 in caps} supported={supported}")
+
+    t0 = time.monotonic()
+    helper.write_packets([(GET_YAML_REQ, b"")], False)
+    chunks: list[bytes] = []
+    total = encoding = None
+    done = False
+    while not done:
+        f = fields(await expect(GET_YAML_RESP, 20))
+        if 1 in f:
+            chunks.append(f[1][0])
+        done = f.get(2) == [1]
+        if 3 in f:
+            total = f[3][0]
+        if 4 in f:
+            encoding = f[4][0].decode()
+    dt = time.monotonic() - t0
+    compressed = b"".join(chunks)
+    print(f"chunks {len(chunks)} sizes {[len(c) for c in chunks]}")
+    print(
+        f"received {len(compressed)} bytes, advertised {total}, encoding {encoding!r}, {dt:.2f} s"
+    )
+    # a second request right after: the device must answer again, not stay stuck
+    helper.write_packets([(GET_YAML_REQ, b"")], False)
+    again = 0
+    done = False
+    while not done:
+        f = fields(await expect(GET_YAML_RESP, 20))
+        again += len(f.get(1, [b""])[0])
+        done = f.get(2) == [1]
+    print(f"second request: {again} bytes")
+    helper.close()
+    if errors:
+        sys.exit(f"transport errors: {errors}")
+    if encoding != "zstd":
+        sys.exit(f"unexpected encoding {encoding!r}, expected 'zstd'")
+    if total is not None and len(compressed) != total:
+        sys.exit(f"received {len(compressed)} bytes but the device advertised {total}")
+    if again != len(compressed):
+        sys.exit(
+            f"the second request returned {again} bytes, the first {len(compressed)}"
+        )
+    envelope = zstd.decompress(compressed)
+    files = unpack_envelope(envelope)
+    print(
+        f"envelope {len(envelope)} bytes, {len(files)} files, ratio {len(compressed) / len(envelope):.2f}"
+    )
+    OUT.mkdir(parents=True, exist_ok=True)
+    for rel, content in files.items():
+        target = OUT / rel.replace("../", "up_/")
+        if not target.resolve().is_relative_to(OUT.resolve()):
+            print("skipping a path outside the output directory:", rel)
+            continue
+        target.parent.mkdir(parents=True, exist_ok=True)
+        target.write_bytes(content)
+        note = ""
+        if ORIG is not None:
+            orig = (ORIG / rel).resolve()
+            if orig.exists():
+                note = (
+                    "IDENTICAL to original"
+                    if orig.read_bytes() == content
+                    else f"DIFFERS from original ({orig.stat().st_size} bytes)"
+                )
+            else:
+                note = "no original"
+        print(f"  {rel}: {len(content)} bytes {note}")
+
+
+asyncio.run(main())

.agents/skills/hardware-test/scripts/http_request/burstonly.py

@@ -0,0 +1,44 @@
+"""Run the burst action once and print each queued request as it completes.
+Usage: burstonly.py IP URL N"""
+
+import asyncio
+import re
+import sys
+import time
+
+import aioesphomeapi
+
+IP, URL, N = sys.argv[1], sys.argv[2], int(sys.argv[3])
+
+
+async def main():
+    c = aioesphomeapi.APIClient(IP, 6053, None)
+    await c.connect(login=True)
+    _, svcs = await c.list_entities_services()
+    s = {x.name: x for x in svcs}
+    t0 = time.monotonic()
+    done = []
+    errors = []
+
+    def on_log(m):
+        t = re.sub(r"\x1b\[[0-9;]*m", "", m.message.decode(errors="replace"))
+        if "BURSTRESP status=200 " in t:
+            done.append(t)
+        elif "BURSTRESP" in t or "BURSTERR" in t:
+            errors.append(t)
+        if "hrt" in t or "Too many" in t:
+            print(f"{time.monotonic() - t0:6.2f} {t[:120]}", flush=True)
+
+    c.subscribe_logs(on_log, log_level=aioesphomeapi.LogLevel.LOG_LEVEL_INFO)
+    await asyncio.sleep(1)
+    t0 = time.monotonic()
+    await c.execute_service(s["burst"], {"url": URL, "count": N})
+    await asyncio.sleep(14)
+    await c.disconnect()
+    print(
+        f"{len(done)} of {N} burst requests answered, {len(errors)} failed", flush=True
+    )
+    return 0 if len(done) >= N else 1
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/http_request/hrdrive.py

@@ -0,0 +1,159 @@
+"""Drive the http_request action test device over the API and collect its hrt log lines.
+Usage: hrdrive.py DEVICE_IP SERVER_BASE LABEL [https]"""
+
+import asyncio
+import os
+import re
+import sys
+import time
+
+import aioesphomeapi
+
+IP, BASE, LABEL = sys.argv[1], sys.argv[2], sys.argv[3]
+HTTPS = len(sys.argv) > 4
+CASES = [
+    ("get", "fast", BASE + "/fast"),
+    ("get", "slow2s", BASE + "/slow?d=2"),
+    ("get", "big70k", BASE + "/big?n=70000&rate=20000"),
+    ("get", "chunked70k", BASE + "/chunked?n=70000&rate=20000"),
+    ("post", "post500", BASE + "/post"),
+    ("get", "redirect", BASE + "/redirect"),
+    ("get", "hang", BASE + "/hang?d=20"),
+    ("get", "unreachable", "http://192.0.2.1:81/"),
+]
+if HTTPS:
+    CASES += [
+        ("get", "https_get", "https://www.howsmyssl.com/a/check"),
+        ("post", "https_post", "https://httpbin.org/post"),
+    ]
+
+ONLY = os.environ.get("ONLY")
+if ONLY:
+    CASES = [c for c in CASES if c[1] in ONLY.split(",")]
+ROUNDS = int(os.environ.get("ROUNDS", "2"))
+# cases whose request must fail (past the 8 s timeout, or an address that never answers)
+EXPECT_ERROR = {"hang", "unreachable"}
+EXTRAS = not ONLY
+
+
+timeouts: list[str] = []
+mismatches: list[str] = []
+
+
+async def main() -> int:
+    c = aioesphomeapi.APIClient(IP, 6053, None)
+    await c.connect(login=True)
+    _, services = await c.list_entities_services()
+    svc = {s.name: s for s in services}
+    lines = []
+
+    def on_log(m):
+        t = (
+            m.message.decode(errors="replace")
+            if isinstance(m.message, bytes)
+            else m.message
+        )
+        t = re.sub(r"\x1b\[[0-9;]*m", "", t)
+        if (
+            "[hrt" in t
+            or "http_request" in t
+            or "Script" in t
+            or "already running" in t
+        ):
+            lines.append(t)
+            print(f"{LABEL} {t}", flush=True)
+
+    def bad_burst(ln: str) -> bool:
+        return "BURSTERR" in ln or ("BURSTRESP" in ln and "status=200 " not in ln)
+
+    def mismatch(what: str, why: str) -> None:
+        print(f"{LABEL} {what} MISMATCH: {why}", flush=True)
+        mismatches.append(what)
+
+    async def wait_until(done, secs: float, what: str) -> None:
+        """Poll done() until it holds; print a TIMEOUT marker when the deadline passes first."""
+        end = time.monotonic() + secs
+        while not done():
+            if time.monotonic() >= end:
+                print(f"{LABEL} {what} TIMEOUT", flush=True)
+                timeouts.append(what)
+                return
+            await asyncio.sleep(0.2)
+
+    c.subscribe_logs(on_log, log_level=aioesphomeapi.LogLevel.LOG_LEVEL_DEBUG)
+    await asyncio.sleep(1)
+    for rnd in range(1, ROUNDS + 1):
+        for kind, name, url in CASES:
+            n0 = len(lines)
+            print(f"{LABEL} == {name} round {rnd}", flush=True)
+            if kind == "get":
+                await c.execute_service(svc["req_get"], {"url": url})
+            else:
+                await c.execute_service(
+                    svc["req_post"], {"url": url, "payload": "x" * 500}
+                )
+            await wait_until(
+                lambda n0=n0: any("RESULT" in ln for ln in lines[n0:]),
+                40,
+                f"{name} round {rnd}",
+            )
+            result = next((ln for ln in lines[n0:] if "RESULT" in ln), None)
+            if result is not None and (
+                ("status=200 " not in result)
+                if name not in EXPECT_ERROR
+                else (" error " not in result)
+            ):
+                mismatch(f"{name} round {rnd}", result)
+            await asyncio.sleep(1.5)
+        if not EXTRAS:
+            continue
+        print(f"{LABEL} == burst round {rnd}", flush=True)
+        n0 = len(lines)
+        await c.execute_service(svc["burst"], {"url": BASE + "/slow?d=1", "count": 6})
+        await wait_until(
+            lambda n0=n0: (
+                sum("BURSTRESP" in ln or "BURSTERR" in ln for ln in lines[n0:]) >= 6
+            ),
+            30,
+            f"burst round {rnd}",
+        )
+        if any(bad_burst(ln) for ln in lines[n0:]):
+            mismatch(f"burst round {rnd}", "a burst request failed or was not 200")
+        await asyncio.sleep(2)
+        print(f"{LABEL} == cancel round {rnd}", flush=True)
+        n0 = len(lines)
+        await c.execute_service(svc["cancel"], {"url": BASE + "/slow?d=2"})
+        await asyncio.sleep(9)
+        if not any("CANCEL stopped running=0" in ln for ln in lines[n0:]):
+            mismatch(
+                f"cancel round {rnd}", "script.stop did not leave the script stopped"
+            )
+        print(f"{LABEL} == cross round {rnd}", flush=True)
+        n0 = len(lines)
+        await c.execute_service(svc["req_get"], {"url": BASE + "/slow?d=2&tag=cross"})
+        await c.execute_service(
+            svc["burst"], {"url": BASE + "/slow?d=1&tag=crossburst", "count": 3}
+        )
+        await wait_until(
+            lambda n0=n0: (
+                sum("BURSTRESP" in ln or "BURSTERR" in ln for ln in lines[n0:]) >= 3
+                and any("RESULT" in ln for ln in lines[n0:])
+            ),
+            30,
+            f"cross round {rnd}",
+        )
+        if any(
+            bad_burst(ln) or ("RESULT" in ln and "status=200 " not in ln)
+            for ln in lines[n0:]
+        ):
+            mismatch(f"cross round {rnd}", "a request failed")
+        await asyncio.sleep(2)
+    await c.disconnect()
+    if timeouts:
+        print(f"{LABEL} TIMEOUTS: {', '.join(timeouts)}", flush=True)
+    if mismatches:
+        print(f"{LABEL} MISMATCHES: {', '.join(mismatches)}", flush=True)
+    return 1 if timeouts or mismatches else 0
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/http_request/hsrv.py

@@ -0,0 +1,78 @@
+"""Test server for http_request actions: deterministic bodies, slow and chunked responses."""
+
+from http.server import BaseHTTPRequestHandler, ThreadingHTTPServer
+import sys
+import time
+from urllib.parse import parse_qs, urlparse
+
+
+def body_bytes(n):
+    return bytes((i * 7 + 3) & 0xFF for i in range(n))
+
+
+class H(BaseHTTPRequestHandler):
+    protocol_version = "HTTP/1.1"
+
+    def log_message(self, fmt, *a):
+        sys.stdout.write(f"{time.time():.3f} {self.client_address[0]} {fmt % a}\n")
+        sys.stdout.flush()
+
+    def _q(self):
+        u = urlparse(self.path)
+        return u.path, {k: v[0] for k, v in parse_qs(u.query).items()}
+
+    def _send(self, data, code=200, ctype="application/octet-stream"):
+        self.send_response(code)
+        self.send_header("Content-Type", ctype)
+        self.send_header("Content-Length", str(len(data)))
+        self.end_headers()
+        self.wfile.write(data)
+
+    def do_POST(self):
+        n = int(self.headers.get("Content-Length", 0))
+        got = self.rfile.read(n)
+        self._send(
+            b"posted %d %d" % (len(got), sum(got) & 0xFFFFFFFF), ctype="text/plain"
+        )
+
+    def do_GET(self):
+        path, q = self._q()
+        if path == "/fast":
+            self._send(b"hello", ctype="text/plain")
+        elif path == "/slow":
+            time.sleep(float(q.get("d", "2")))
+            self._send(b"slow ok", ctype="text/plain")
+        elif path == "/redirect":
+            self.send_response(302)
+            self.send_header("Location", "/fast")
+            self.send_header("Content-Length", "0")
+            self.end_headers()
+        elif path == "/hang":
+            time.sleep(float(q.get("d", "30")))
+        elif path in ("/big", "/chunked"):
+            n = int(q.get("n", "70000"))
+            rate = float(q.get("rate", "20000"))
+            data = body_bytes(n)
+            self.send_response(200)
+            self.send_header("Content-Type", "application/octet-stream")
+            if path == "/big":
+                self.send_header("Content-Length", str(n))
+            else:
+                self.send_header("Transfer-Encoding", "chunked")
+            self.end_headers()
+            step = 1000
+            for i in range(0, n, step):
+                part = data[i : i + step]
+                if path == "/big":
+                    self.wfile.write(part)
+                else:
+                    self.wfile.write(b"%x\r\n" % len(part) + part + b"\r\n")
+                self.wfile.flush()
+                time.sleep(len(part) / rate)
+            if path == "/chunked":
+                self.wfile.write(b"0\r\n\r\n")
+        else:
+            self._send(b"nope", 404)
+
+
+ThreadingHTTPServer(("0.0.0.0", int(sys.argv[1])), H).serve_forever()

.agents/skills/hardware-test/scripts/http_request/manifest_server.py

@@ -0,0 +1,58 @@
+"""Serve update manifests: stable.json, beta.json and slow.json (answered after a delay); logs every request."""
+
+from http.server import BaseHTTPRequestHandler, ThreadingHTTPServer
+import json
+import sys
+import time
+
+PORT = int(sys.argv[1])
+DELAY = float(sys.argv[2]) if len(sys.argv) > 2 else 4.0
+
+
+def manifest(name: str, version: str) -> bytes:
+    builds = [
+        {
+            "chipFamily": fam,
+            "ota": {
+                "path": f"/{name}.bin",
+                "md5": "d41d8cd98f00b204e9800998ecf8427e",
+                "summary": f"{name} channel",
+            },
+        }
+        for fam in ("ESP8266", "ESP32")
+    ]
+    return json.dumps(
+        {"name": f"Bench {name}", "version": version, "builds": builds}
+    ).encode()
+
+
+DOCS = {
+    "/stable.json": manifest("stable", "1.0.0"),
+    "/beta.json": manifest("beta", "2.0.0-beta1"),
+    "/slow.json": manifest("slow", "3.0.0-slow"),
+}
+
+
+class Handler(BaseHTTPRequestHandler):
+    def do_GET(self) -> None:  # noqa: N802
+        body = DOCS.get(self.path)
+        print(
+            f"{time.strftime('%H:%M:%S')} {self.client_address[0]} GET {self.path} -> {404 if body is None else 200}",
+            flush=True,
+        )
+        if body is None:
+            self.send_error(404)
+            return
+        if self.path == "/slow.json":
+            time.sleep(DELAY)
+        self.send_response(200)
+        self.send_header("Content-Type", "application/json")
+        self.send_header("Content-Length", str(len(body)))
+        self.end_headers()
+        self.wfile.write(body)
+
+    def log_message(self, *args) -> None:  # quiet default log
+        return
+
+
+ThreadingHTTPServer(("0.0.0.0", PORT), Handler).serve_forever()

.agents/skills/hardware-test/scripts/http_request/ota_auth/base/32/ho32base.yaml

@@ -0,0 +1,10 @@
+esphome:
+  name: ho32base
+
+esp32:
+  board: esp32dev
+  framework:
+    type: esp-idf
+
+packages:
+  common: !include ../../ho-common.yaml

.agents/skills/hardware-test/scripts/http_request/ota_auth/base/66/ho66base.yaml

@@ -0,0 +1,11 @@
+esphome:
+  name: ho66base
+
+esp8266:
+  board: d1_mini
+
+packages:
+  common: !include ../../ho-common.yaml
+
+http_request:
+  verify_ssl: false

.agents/skills/hardware-test/scripts/http_request/ota_auth/fwsrv.py

@@ -0,0 +1,32 @@
+"""Serve a www dir for ota.http_request tests; /auth/ needs basic auth otauser:otapass.
+Usage: fwsrv.py WWW PORT"""
+
+import base64
+import functools
+import http.server
+import sys
+
+WWW, PORT = sys.argv[1], int(sys.argv[2])
+GOOD = "Basic " + base64.b64encode(b"otauser:otapass").decode()
+
+
+class H(http.server.SimpleHTTPRequestHandler):
+    def do_GET(self):
+        if self.path.startswith("/auth/") and self.headers.get("Authorization") != GOOD:
+            self.send_response(401)
+            self.send_header("WWW-Authenticate", 'Basic realm="ota"')
+            self.send_header("Content-Length", "0")
+            self.end_headers()
+            return
+        super().do_GET()
+
+    def log_message(self, fmt, *args):
+        print(
+            f"{self.address_string()} auth={'Authorization' in self.headers} {fmt % args}",
+            flush=True,
+        )
+
+
+http.server.ThreadingHTTPServer(
+    ("0.0.0.0", PORT), functools.partial(H, directory=WWW)
+).serve_forever()

.agents/skills/hardware-test/scripts/http_request/ota_auth/ho-common.yaml

@@ -0,0 +1,60 @@
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+logger:
+  level: DEBUG
+
+api:
+  reboot_timeout: 0s
+  actions:
+    - action: fl_md5url
+      variables: {url: string, md5url: string}
+      then:
+        - ota.http_request.flash:
+            url: !lambda return url;
+            md5_url: !lambda return md5url;
+    - action: fl_md5
+      variables: {url: string, sum: string}
+      then:
+        - ota.http_request.flash:
+            url: !lambda return url;
+            md5: !lambda return sum;
+    - action: fl_auth
+      variables: {url: string, md5url: string, user: string, pw: string}
+      then:
+        - ota.http_request.flash:
+            url: !lambda return url;
+            md5_url: !lambda return md5url;
+            username: !lambda return user;
+            password: !lambda return pw;
+    - action: fl_nourl
+      then:
+        - lambda: id(hota).flash();
+    - action: fl_badurl
+      then:
+        - lambda: |-
+            id(hota).set_url("not a url");
+            id(hota).flash();
+    - action: getters
+      then:
+        - lambda: |-
+            ESP_LOGI("hot", "getters computed='%s' expected='%s'", id(hota).md5_computed().c_str(),
+                     id(hota).md5_expected().c_str());
+
+esphome:
+  on_boot:
+    priority: -100
+    then:
+      - lambda: |-
+          ESP_LOGI("hot", "boot getters computed='%s' expected='%s'", id(hota).md5_computed().c_str(),
+                   id(hota).md5_expected().c_str());
+
+http_request:
+  timeout: 10s
+
+ota:
+  - platform: esphome
+  - platform: http_request
+    id: hota

.agents/skills/hardware-test/scripts/http_request/ota_auth/hodrive.py

@@ -0,0 +1,142 @@
+"""Drive the ota.http_request cases over the API.
+Usage: hodrive.py IP BASEURL MD5FILE (BASEURL serves fw.bin, fw.md5, bad.md5, auth/fw.bin, auth/fw.md5)"""
+
+import asyncio
+import pathlib
+import re
+import sys
+import time
+
+import aioesphomeapi
+
+if len(sys.argv) != 4:
+    sys.exit(__doc__)
+IP, BASE = sys.argv[1:3]
+SUM = pathlib.Path(sys.argv[3]).read_text(encoding="utf-8").strip()
+T0 = time.monotonic()
+lines: list[str] = []
+failed: list[str] = []
+
+
+def out(s: str) -> None:
+    print(f"{time.monotonic() - T0:7.2f} {s}", flush=True)
+
+
+async def connect() -> tuple[aioesphomeapi.APIClient, dict]:
+    last: Exception | None = None
+    for _ in range(90):
+        c = aioesphomeapi.APIClient(IP, 6053, None)
+        try:
+            await c.connect(login=True)
+        except (
+            aioesphomeapi.InvalidAuthAPIError,
+            aioesphomeapi.InvalidEncryptionKeyAPIError,
+            aioesphomeapi.RequiresEncryptionAPIError,
+        ):
+            raise  # a key or password problem never fixes itself
+        except Exception as exc:  # noqa: BLE001
+            last = exc
+            await asyncio.sleep(2)
+            continue
+        i = await c.device_info()
+        out(f"connected {i.name} {i.compilation_time}")
+        _, svcs = await c.list_entities_services()
+
+        def on_log(m):
+            t = re.sub(r"\x1b\[[0-9;]*m", "", m.message.decode(errors="replace"))
+            if re.search(r"\[hot|http_request|ota|OTA|MD5|md5", t):
+                lines.append(t)
+                out(t[:170])
+
+        c.subscribe_logs(on_log, log_level=aioesphomeapi.LogLevel.LOG_LEVEL_DEBUG)
+        return c, {x.name: x for x in svcs}
+    raise SystemExit(f"device never came back (last error {last!r})")
+
+
+async def case(c, s, name, args, want, timeout=60.0):
+    out(f"=== {name} {args} (want {want})")
+    start = len(lines)
+    await c.execute_service(s[name], args)
+    end = time.monotonic() + timeout
+    while time.monotonic() < end:
+        if any(re.search(want, x) for x in lines[start:]):
+            out(f"RESULT {name} PASS")
+            return True
+        await asyncio.sleep(0.2)
+    out(f"RESULT {name} FAIL (no '{want}')")
+    failed.append(name)
+    return False
+
+
+async def flash_ok(c, s, name, args):
+    """A good flash: wait for the device to drop and come back."""
+    out(f"=== {name} {args} (want reboot)")
+    start = len(lines)
+    await c.execute_service(s[name], args)
+    end = time.monotonic() + 120
+    while (
+        time.monotonic() < end
+        and c._connection is not None
+        and c._connection.is_connected
+    ):  # noqa: SLF001
+        await asyncio.sleep(0.5)
+    # a good flash needs both the device's own success line and the connection dropping for the reboot
+    dropped = c._connection is None or not c._connection.is_connected  # noqa: SLF001
+    logged = any(
+        re.search(r"(?i)update complete|rebooting|successful", x) for x in lines[start:]
+    )
+    if not dropped:
+        out(f"{name}: the device never dropped the connection within 120 s")
+    ok = logged and dropped
+    await c.disconnect()
+    await asyncio.sleep(5)
+    c, s = await connect()
+    out(f"RESULT {name} {'PASS' if ok else 'FAIL'}")
+    if not ok:
+        failed.append(name)
+    return c, s
+
+
+async def main():
+    c, s = await connect()
+    u, m = f"{BASE}/fw.bin", f"{BASE}/fw.md5"
+    await case(c, s, "getters", {}, r"computed='' expected=''")
+    await case(c, s, "fl_nourl", {}, r"URL not set")
+    await case(c, s, "fl_badurl", {}, r"URL not set")
+    await case(c, s, "getters", {}, r"computed='' expected=''")
+    await case(c, s, "fl_md5", {"url": u, "sum": "0" * 32}, r"MD5 mismatch")
+    await case(c, s, "getters", {}, r"getters computed=")
+    await case(
+        c,
+        s,
+        "fl_auth",
+        {
+            "url": f"{BASE}/auth/fw.bin",
+            "md5url": f"{BASE}/auth/fw.md5",
+            "user": "otauser",
+            "pw": "wrong",
+        },
+        r"HTTP Request failed; URL: \S*/auth/fw\.md5; Code: 401",
+    )
+    c, s = await flash_ok(
+        c,
+        s,
+        "fl_auth",
+        {
+            "url": f"{BASE}/auth/fw.bin",
+            "md5url": f"{BASE}/auth/fw.md5",
+            "user": "otauser",
+            "pw": "otapass",
+        },
+    )
+    c, s = await flash_ok(c, s, "fl_md5url", {"url": u, "md5url": m})
+    c, s = await flash_ok(c, s, "fl_md5", {"url": u, "sum": SUM})
+    await case(c, s, "getters", {}, r"computed='' expected=''")
+    await c.disconnect()
+    out(
+        f"SUMMARY {'all cases passed' if not failed else 'FAILED: ' + ', '.join(failed)}"
+    )
+    return 1 if failed else 0
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/http_request/ota_auth/pr/32/ho32pr.yaml

@@ -0,0 +1,10 @@
+esphome:
+  name: ho32pr
+
+esp32:
+  board: esp32dev
+  framework:
+    type: esp-idf
+
+packages:
+  common: !include ../../ho-common.yaml

.agents/skills/hardware-test/scripts/http_request/ota_auth/pr/66/ho66pr.yaml

@@ -0,0 +1,11 @@
+esphome:
+  name: ho66pr
+
+esp8266:
+  board: d1_mini
+
+packages:
+  common: !include ../../ho-common.yaml
+
+http_request:
+  verify_ssl: false

.agents/skills/hardware-test/scripts/http_request/ota_auth/run.sh

@@ -0,0 +1,30 @@
+#!/bin/bash
+# run.sh BUILD(pr|base) BOARD(32|66) : serial flash ho<BOARD><BUILD>, serve its own image, drive the cases
+# Copy this directory to a scratch directory (secrets.yaml next to ho-common.yaml) and run it there.
+# env: TREE (esphome checkout for BUILD), P32 or P66 (serial port of the board), SC (path to serial_capture.py),
+# HOST_IP (this host's address as the device reaches it), PY (checkout python)
+D=$(cd "$(dirname "$0")" && pwd); V=$1; B=$2; N=ho$B$V; T=${TREE:?set TREE}; PY=${PY:-$T/venv/bin/python}
+SC=${SC:?set SC}; HOST=${HOST_IP:?set HOST_IP}
+if [ "$B" = 32 ]; then P=${P32:?set P32}; else P=${P66:?set P66}; fi
+OUT=$D/out-$V-$B; rm -rf "$OUT"; mkdir -p "$OUT"
+pkill -f "[s]erial_capture.py $P"
+(cd "$T" && $PY -m esphome compile "$D/$V/$B/$N.yaml" > "$OUT/compile.log" 2>&1) || { echo "compile failed"; exit 1; }
+(cd "$T" && $PY -m esphome upload "$D/$V/$B/$N.yaml" --device "$P" > "$OUT/upload.log" 2>&1) || { echo "upload failed"; exit 1; }
+# the exact image of this config, just compiled, never one found elsewhere
+if [ "$B" = 32 ]; then IMG=$D/$V/$B/.esphome/build/$N/build/firmware.ota.bin
+else IMG=$D/$V/$B/.esphome/build/$N/.pioenvs/$N/firmware.bin; fi
+[ -f "$IMG" ] || { echo "no image for $N at $IMG"; exit 1; }
+echo "image $IMG"
+W=$OUT/www; mkdir -p "$W/auth"; cp "$IMG" "$W/fw.bin"; cp "$IMG" "$W/auth/fw.bin"
+$PY -c "import hashlib,sys;print(hashlib.md5(open(sys.argv[1],'rb').read()).hexdigest())" "$IMG" > "$W/fw.md5"
+cp "$W/fw.md5" "$W/auth/fw.md5"; echo 00000000000000000000000000000000 > "$W/bad.md5"
+$PY "$SC" "$P" 400 "$OUT/serial.txt" --reset > /dev/null 2>&1 & CAP=$!
+trap 'kill $CAP $SRV 2>/dev/null' EXIT
+pkill -f "[f]wsrv.py"; $PY "$D/fwsrv.py" "$W" 8093 > "$OUT/server.txt" 2>&1 & SRV=$!
+for i in $(seq 60); do IP=$(grep -a -o 'IP Address: [0-9.]*' "$OUT/serial.txt" | tail -1 | cut -d' ' -f3); [ -z "$IP" ] && IP=$(python3 -c "import socket;print(socket.gethostbyname('$N.local'))" 2>/dev/null); [ -n "$IP" ] && break; sleep 2; done
+[ -n "$IP" ] || { echo "device not found"; exit 1; }
+echo "device up"
+$PY "$D/hodrive.py" "$IP" "http://$HOST:8093" "$W/fw.md5" > "$OUT/drive.txt" 2>&1; RC=$?
+kill $SRV; pkill -f "[s]erial_capture.py $P"
+grep -E 'RESULT|SUMMARY|never came back' "$OUT/drive.txt"; echo "server requests: $(grep -c . "$OUT/server.txt")"
+exit $RC

.agents/skills/hardware-test/scripts/http_request/otadrive.py

@@ -0,0 +1,74 @@
+"""Trigger ota.http_request.flash over the API, optionally during a burst of action requests.
+Usage: otadrive.py IP FWURL MD5 BASE [burst]"""
+
+import asyncio
+import contextlib
+import re
+import sys
+import time
+
+import aioesphomeapi
+
+IP, FW, SUM, BASE = sys.argv[1:5]
+BURST = len(sys.argv) > 5
+
+
+async def main():
+    c = aioesphomeapi.APIClient(IP, 6053, None)
+    await c.connect(login=True)
+    i = await c.device_info()
+    print("before:", i.name, i.compilation_time, flush=True)
+    _, svcs = await c.list_entities_services()
+    s = {x.name: x for x in svcs}
+    t0 = time.monotonic()
+    seen: list[str] = []
+
+    def on_log(m):
+        t = re.sub(r"\x1b\[[0-9;]*m", "", m.message.decode(errors="replace"))
+        seen.append(t)
+        if re.search(r"hrt|http_request|ota|OTA|Too many", t):
+            print(f"{time.monotonic() - t0:7.2f} {t[:170]}", flush=True)
+
+    c.subscribe_logs(on_log, log_level=aioesphomeapi.LogLevel.LOG_LEVEL_INFO)
+    await asyncio.sleep(1)
+    if BURST:
+        await c.execute_service(s["burst"], {"url": BASE + "/slow?d=1", "count": 6})
+        await asyncio.sleep(0.5)
+    await c.execute_service(s["otaflash"], {"url": FW, "sum": SUM})
+    end = time.monotonic() + 45
+    while (
+        time.monotonic() < end
+        and c._connection is not None  # noqa: SLF001
+        and c._connection.is_connected  # noqa: SLF001
+    ):
+        await asyncio.sleep(0.5)
+    dropped = c._connection is None or not c._connection.is_connected  # noqa: SLF001
+    with contextlib.suppress(Exception):
+        await c.disconnect()
+    for _ in range(30):
+        try:
+            c2 = aioesphomeapi.APIClient(IP, 6053, None)
+            await c2.connect(login=True)
+            i = await c2.device_info()
+            print("after:", i.name, i.compilation_time, flush=True)
+            await c2.disconnect()
+            break
+        except Exception:  # noqa: BLE001
+            await asyncio.sleep(2)
+    else:
+        print("RESULT FAIL: not reachable after the flash", flush=True)
+        return 1
+    if not dropped:
+        print("RESULT FAIL: the device never rebooted", flush=True)
+        return 1
+    if any("OTAFLASH returned" in t for t in seen):
+        print("RESULT FAIL: the flash action returned, the update failed", flush=True)
+        return 1
+    if not any(re.search(r"(?i)update complete|rebooting|successful", t) for t in seen):
+        print("RESULT FAIL: no update complete line before the reboot", flush=True)
+        return 1
+    print("RESULT PASS", flush=True)
+    return 0
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/http_request/reboottest.py

@@ -0,0 +1,68 @@
+"""Call reboot_during (a slow request in flight, another waiting, then App.safe_reboot) and time the reboot.
+Usage: reboottest.py IP URL"""
+
+import asyncio
+import contextlib
+import re
+import sys
+import time
+
+import aioesphomeapi
+
+IP, URL = sys.argv[1], sys.argv[2]
+
+
+async def main():
+    c = aioesphomeapi.APIClient(IP, 6053, None)
+    await c.connect(login=True)
+    _, svcs = await c.list_entities_services()
+    s = {x.name: x for x in svcs}
+    t0 = time.monotonic()
+
+    def on_log(m):
+        t = re.sub(r"\x1b\[[0-9;]*m", "", m.message.decode(errors="replace"))
+        if re.search(
+            r"hrt|http_request|safe_mode|shutdown|Rebooting|reboot", t, re.IGNORECASE
+        ):
+            print(f"{time.monotonic() - t0:7.2f} {t[:150]}", flush=True)
+
+    c.subscribe_logs(on_log, log_level=aioesphomeapi.LogLevel.LOG_LEVEL_DEBUG)
+    await asyncio.sleep(1)
+    t0 = time.monotonic()
+    await c.execute_service(s["reboot_during"], {"url": URL})
+    dropped = False
+    while time.monotonic() - t0 < 30:
+        try:
+            await asyncio.wait_for(c.device_info(), 1.0)
+            await asyncio.sleep(0.2)
+        except TimeoutError:
+            continue  # a slow reply is not a reboot; only a closed connection counts
+        except Exception:  # noqa: BLE001
+            dropped = True
+            print(f"{time.monotonic() - t0:7.2f} API connection gone", flush=True)
+            break
+    with contextlib.suppress(Exception):
+        await c.disconnect()
+    if not dropped:
+        print(f"{time.monotonic() - t0:7.2f} NO REBOOT within 30 s", flush=True)
+        return 1
+    last: Exception | None = None
+    for _ in range(60):
+        try:
+            c2 = aioesphomeapi.APIClient(IP, 6053, None)
+            await asyncio.wait_for(c2.connect(login=True), 3)
+            i = await c2.device_info()
+            print(f"{time.monotonic() - t0:7.2f} back: {i.name}", flush=True)
+            await c2.disconnect()
+            return 0
+        except Exception as exc:  # noqa: BLE001
+            last = exc
+            await asyncio.sleep(1)
+    print(
+        f"{time.monotonic() - t0:7.2f} NOT BACK after reboot (last error {last!r})",
+        flush=True,
+    )
+    return 1
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/http_request/updcheck.py

@@ -0,0 +1,66 @@
+"""Run N update checks over the API and print the update state plus http_request/update/long-operation log lines.
+Usage: updcheck.py IP LABEL N"""
+
+import asyncio
+import re
+import sys
+import time
+
+import aioesphomeapi
+
+IP, LABEL, N = sys.argv[1], sys.argv[2], int(sys.argv[3])
+
+
+async def main():
+    c = aioesphomeapi.APIClient(IP, 6053, None)
+    await c.connect(login=True)
+    info = await c.device_info()
+    print(
+        f"{LABEL} device {info.name} {info.esphome_version} {info.compilation_time}",
+        flush=True,
+    )
+    ents, _ = await c.list_entities_services()
+    upd = next(e for e in ents if isinstance(e, aioesphomeapi.UpdateInfo))
+    states = []
+
+    def on_state(s):
+        if s.key == upd.key:
+            states.append(s)
+            print(
+                f"{LABEL} {time.monotonic() - t0:7.2f} UPDATE in_progress={s.in_progress} current={s.current_version} latest={s.latest_version} title={s.title!r}",
+                flush=True,
+            )
+
+    def on_log(m):
+        t = (
+            m.message.decode(errors="replace")
+            if isinstance(m.message, bytes)
+            else m.message
+        )
+        t = re.sub(r"\x1b\[[0-9;]*m", "", t)
+        if re.search(
+            r"http_request|update|took a long time|Guru|abort|heap|Free",
+            t,
+            re.IGNORECASE,
+        ):
+            print(f"{LABEL} {time.monotonic() - t0:7.2f} {t[:200]}", flush=True)
+
+    t0 = time.monotonic()
+    c.subscribe_states(on_state)
+    c.subscribe_logs(on_log, log_level=aioesphomeapi.LogLevel.LOG_LEVEL_DEBUG)
+    await asyncio.sleep(3)
+    missing = []
+    for i in range(N):
+        print(f"{LABEL} {time.monotonic() - t0:7.2f} CHECK {i + 1}", flush=True)
+        before = len(states)
+        c.update_command(upd.key, aioesphomeapi.UpdateCommand.CHECK)
+        await asyncio.sleep(25)
+        if not any(not st.in_progress for st in states[before:]):
+            print(f"{LABEL} CHECK {i + 1} got no finished update state", flush=True)
+            missing.append(i + 1)
+    await c.disconnect()
+    print(f"{LABEL} {N - len(missing)} of {N} checks reported a state", flush=True)
+    return 1 if missing else 0
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/http_request/updrive.py

@@ -0,0 +1,129 @@
+"""Drive the http_request update harness over the API: press buttons, print every update state and save the log.
+usage: updrive.py HOST KEY(- for plaintext) LOGFILE OP [OP ...]
+  OP: list | press:NAME | set:NAME=VALUE | sel:NAME=OPTION | wait:SECONDS
+"""
+
+import asyncio
+import pathlib
+import sys
+import time
+
+from aioesphomeapi import (
+    APIClient,
+    BinarySensorState,
+    ButtonInfo,
+    LogLevel,
+    NumberInfo,
+    NumberState,
+    SelectInfo,
+    SelectState,
+    SensorInfo,
+    SensorState,
+    TextSensorInfo,
+    TextSensorState,
+    UpdateState,
+)
+
+HOST, KEY, LOGFILE = sys.argv[1], sys.argv[2], sys.argv[3]
+OPS = sys.argv[4:]
+T0 = time.time()
+
+
+def stamp() -> str:
+    return f"{time.time() - T0:7.3f}"
+
+
+async def main() -> None:
+    with pathlib.Path(LOGFILE).open("w", encoding="utf-8") as log:
+        c = APIClient(HOST, 6053, None, noise_psk=None if KEY == "-" else KEY)
+        await c.connect(login=True)
+        info = await c.device_info()
+        print(
+            f"device {info.name} esphome {info.esphome_version} compiled {info.compilation_time}"
+        )
+        ents, _ = await c.list_entities_services()
+        by_key = {e.key: e for e in ents}
+        buttons = {e.name: e for e in ents if isinstance(e, ButtonInfo)}
+        selects = {e.name: e for e in ents if isinstance(e, SelectInfo)}
+
+        def on_state(s) -> None:
+            e = by_key.get(s.key)
+            if e is None:
+                return
+            if isinstance(s, TextSensorState):
+                print(
+                    f"{stamp()} TXT {e.name} = {s.state!r}{' (missing)' if s.missing_state else ''}",
+                    flush=True,
+                )
+            elif isinstance(s, UpdateState):
+                print(
+                    f"{stamp()} UPD {e.name} latest={s.latest_version!r} title={s.title!r} summary={s.release_summary!r} in_progress={s.in_progress}{' (missing)' if s.missing_state else ''}",
+                    flush=True,
+                )
+            elif isinstance(s, BinarySensorState):
+                print(
+                    f"{stamp()} BIN {e.name} = {s.state}{' (missing)' if s.missing_state else ''}",
+                    flush=True,
+                )
+            elif isinstance(s, NumberState):
+                print(
+                    f"{stamp()} NUM {e.name} = {s.state:.0f}{' (missing)' if s.missing_state else ''}",
+                    flush=True,
+                )
+            elif isinstance(s, SelectState):
+                print(
+                    f"{stamp()} SEL {e.name} = {s.state!r}{' (missing)' if s.missing_state else ''}",
+                    flush=True,
+                )
+            elif isinstance(s, SensorState) and isinstance(e, SensorInfo):
+                print(f"{stamp()} SEN {e.name} = {s.state:.0f}", flush=True)
+
+        def on_log(msg) -> None:
+            log.write(f"{stamp()} {msg.message.decode(errors='replace')}\n")
+
+        c.subscribe_states(on_state)
+        c.subscribe_logs(on_log, log_level=LogLevel.LOG_LEVEL_VERBOSE)
+        await asyncio.sleep(1.5)
+        for op in OPS:
+            if op == "list":
+                for n in sorted(buttons):
+                    print("button:", n)
+                for n in sorted(selects):
+                    print("select:", n, selects[n].options)
+                for e in ents:
+                    if isinstance(e, TextSensorInfo):
+                        print("text_sensor:", e.name)
+                continue
+            kind, _, rest = op.partition(":")
+            if kind == "wait":
+                await asyncio.sleep(float(rest))
+                continue
+            if kind == "press":
+                if rest not in buttons:
+                    sys.exit(f"no button named {rest!r}; have {sorted(buttons)}")
+                print(f"{stamp()} >>> press {rest}", flush=True)
+                log.write(f"{stamp()} >>> press {rest}\n")
+                c.button_command(buttons[rest].key)
+            elif kind == "set":
+                name, _, value = rest.partition("=")
+                numbers = {e.name: e for e in ents if isinstance(e, NumberInfo)}
+                if name not in numbers:
+                    sys.exit(f"no number named {name!r}")
+                print(f"{stamp()} >>> set {name} = {value}", flush=True)
+                log.write(f"{stamp()} >>> set {name} = {value}\n")
+                c.number_command(numbers[name].key, float(value))
+            elif kind == "sel":
+                name, _, value = rest.partition("=")
+                if name not in selects:
+                    sys.exit(f"no select named {name!r}; have {sorted(selects)}")
+                print(f"{stamp()} >>> select {name} = {value}", flush=True)
+                log.write(f"{stamp()} >>> select {name} = {value}\n")
+                c.select_command(selects[name].key, value)
+            else:
+                sys.exit(f"unknown op {op!r}")
+            await asyncio.sleep(2.0)
+        await asyncio.sleep(1)
+        await c.disconnect()
+
+
+asyncio.run(main())

.agents/skills/hardware-test/scripts/modem/mdrive.py

@@ -0,0 +1,86 @@
+"""Drive the modem test build over the API: press buttons, print sensor states, save the device log.
+usage: mdrive.py HOST LOGFILE OP [OP ...]
+  OP: list | press:NAME | wait:SECONDS
+"""
+
+import asyncio
+import pathlib
+import sys
+import time
+
+from aioesphomeapi import (
+    APIClient,
+    ButtonInfo,
+    LogLevel,
+    SensorInfo,
+    SensorState,
+    TextSensorInfo,
+    TextSensorState,
+)
+
+HOST, LOGFILE = sys.argv[1], sys.argv[2]
+OPS = sys.argv[3:]
+T0 = time.time()
+
+
+def stamp() -> str:
+    return f"{time.time() - T0:7.3f}"
+
+
+async def main() -> None:
+    with pathlib.Path(LOGFILE).open("w", encoding="utf-8") as log:
+        c = APIClient(HOST, 6053, None)
+        await c.connect(login=True)
+        info = await c.device_info()
+        print(
+            f"device {info.name} esphome {info.esphome_version} compiled {info.compilation_time}"
+        )
+        ents, _ = await c.list_entities_services()
+        by_key = {e.key: e for e in ents}
+        buttons = {e.name: e for e in ents if isinstance(e, ButtonInfo)}
+
+        def on_state(s) -> None:
+            e = by_key.get(s.key)
+            if e is None:
+                return
+            if isinstance(s, TextSensorState):
+                print(
+                    f"{stamp()} TXT {e.name} = {s.state!r}{' (missing)' if s.missing_state else ''}",
+                    flush=True,
+                )
+            elif isinstance(s, SensorState) and isinstance(e, SensorInfo):
+                print(f"{stamp()} SEN {e.name} = {s.state:.1f}", flush=True)
+
+        def on_log(msg) -> None:
+            log.write(f"{stamp()} {msg.message.decode(errors='replace')}\n")
+            log.flush()
+
+        c.subscribe_states(on_state)
+        c.subscribe_logs(on_log, log_level=LogLevel.LOG_LEVEL_DEBUG)
+        await asyncio.sleep(1.5)
+        for op in OPS:
+            if op == "list":
+                for n in sorted(buttons):
+                    print("button:", n)
+                for e in ents:
+                    if isinstance(e, (TextSensorInfo, SensorInfo)):
+                        print("entity:", type(e).__name__, e.name)
+                continue
+            kind, _, rest = op.partition(":")
+            if kind == "wait":
+                await asyncio.sleep(float(rest))
+                continue
+            if kind == "press":
+                if rest not in buttons:
+                    sys.exit(f"no button named {rest!r}; have {sorted(buttons)}")
+                print(f"{stamp()} >>> press {rest}", flush=True)
+                log.write(f"{stamp()} >>> press {rest}\n")
+                c.button_command(buttons[rest].key)
+            else:
+                sys.exit(f"unknown op {op!r}")
+            await asyncio.sleep(2.0)
+        await asyncio.sleep(1)
+        await c.disconnect()
+
+
+asyncio.run(main())

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/base/32/mqa32base.yaml

@@ -0,0 +1,8 @@
+substitutions: {name: mqabase, suffix: "true"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/base/32/mqb32base.yaml

@@ -0,0 +1,9 @@
+substitutions: {name: mqbbase, suffix: "true"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s
+  topic_prefix: custpre32

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/base/32/mqc32base.yaml

@@ -0,0 +1,18 @@
+substitutions: {name: mqcbase, suffix: "true"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s
+  log_topic: explicit32/log
+  birth_message:
+    topic: explicit32/avail
+    payload: up
+  will_message:
+    topic: explicit32/avail
+    payload: down
+  shutdown_message:
+    topic: explicit32/avail
+    payload: bye

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/base/32/mqd32base.yaml

@@ -0,0 +1,8 @@
+substitutions: {name: mqd32base, suffix: "false"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/base/66/mqa66base.yaml

@@ -0,0 +1,7 @@
+substitutions: {name: mqabase, suffix: "true"}
+esp8266:
+  board: d1_mini
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/base/66/mqb66base.yaml

@@ -0,0 +1,8 @@
+substitutions: {name: mqbbase, suffix: "true"}
+esp8266:
+  board: d1_mini
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s
+  topic_prefix: custpre66

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/base/66/mqc66base.yaml

@@ -0,0 +1,17 @@
+substitutions: {name: mqcbase, suffix: "true"}
+esp8266:
+  board: d1_mini
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s
+  log_topic: explicit66/log
+  birth_message:
+    topic: explicit66/avail
+    payload: up
+  will_message:
+    topic: explicit66/avail
+    payload: down
+  shutdown_message:
+    topic: explicit66/avail
+    payload: bye

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/base/66/mqd66base.yaml

@@ -0,0 +1,7 @@
+substitutions: {name: mqd66base, suffix: "false"}
+esp8266:
+  board: d1_mini
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/mq-common.yaml

@@ -0,0 +1,24 @@
+# Package for esphome#9315, included by the per-variant files under pr/ and base/ (substitutions name, suffix).
+esphome:
+  name: ${name}
+  name_add_mac_suffix: ${suffix}
+
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+logger:
+  level: DEBUG
+
+ota:
+  - platform: esphome
+
+button:
+  - platform: restart
+    name: Restart
+
+interval:
+  - interval: 10s
+    then:
+      - logger.log: "alive"

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/pr/32/mqa32pr.yaml

@@ -0,0 +1,8 @@
+substitutions: {name: mqapr, suffix: "true"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/pr/32/mqb32pr.yaml

@@ -0,0 +1,9 @@
+substitutions: {name: mqbpr, suffix: "true"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s
+  topic_prefix: custpre32

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/pr/32/mqc32pr.yaml

@@ -0,0 +1,18 @@
+substitutions: {name: mqcpr, suffix: "true"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s
+  log_topic: explicit32/log
+  birth_message:
+    topic: explicit32/avail
+    payload: up
+  will_message:
+    topic: explicit32/avail
+    payload: down
+  shutdown_message:
+    topic: explicit32/avail
+    payload: bye

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/pr/32/mqd32pr.yaml

@@ -0,0 +1,8 @@
+substitutions: {name: mqd32pr, suffix: "false"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/pr/66/mqa66pr.yaml

@@ -0,0 +1,7 @@
+substitutions: {name: mqapr, suffix: "true"}
+esp8266:
+  board: d1_mini
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/pr/66/mqb66pr.yaml

@@ -0,0 +1,8 @@
+substitutions: {name: mqbpr, suffix: "true"}
+esp8266:
+  board: d1_mini
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s
+  topic_prefix: custpre66

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/pr/66/mqc66pr.yaml

@@ -0,0 +1,17 @@
+substitutions: {name: mqcpr, suffix: "true"}
+esp8266:
+  board: d1_mini
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s
+  log_topic: explicit66/log
+  birth_message:
+    topic: explicit66/avail
+    payload: up
+  will_message:
+    topic: explicit66/avail
+    payload: down
+  shutdown_message:
+    topic: explicit66/avail
+    payload: bye

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/pr/66/mqd66pr.yaml

@@ -0,0 +1,7 @@
+substitutions: {name: mqd66pr, suffix: "false"}
+esp8266:
+  board: d1_mini
+packages: {mq: !include ../../mq-common.yaml}
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s

.agents/skills/hardware-test/scripts/mqtt/mac_suffix/run.sh

@@ -0,0 +1,51 @@
+#!/bin/bash
+# run.sh X BUILD : flash variant X (a..d) of BUILD (pr|base) on both boards, record broker traffic, restart, hard reset
+# Copy this directory to a scratch directory and run it there; secrets.yaml (wifi_ssid, wifi_password, broker_ip)
+# goes next to mq-common.yaml and in every board folder.
+# env: TREE (esphome checkout for BUILD), P32 and P66 (ESP32 and ESP8266 serial ports), SC (path to
+# serial_capture.py), PY (checkout python)
+M=$(cd "$(dirname "$0")" && pwd); X=$1; V=$2; T=${TREE:?set TREE to the checkout under test}
+P32=${P32:?set P32}; P66=${P66:?set P66}; SC=${SC:?set SC}
+PY=${PY:-$T/venv/bin/python}; OUT=$M/out-$X-$V; mkdir -p $OUT
+pkill -f "[s]erial_capture.py $P32"; pkill -f "[s]erial_capture.py $P66"; pkill -f "[m]osquitto_sub -h 127.0.0.1"
+# compile one at a time (shared tool caches), then flash both boards at once; upload alone never builds
+for b in 32 66; do
+  (cd $T && $PY -m esphome compile $M/$V/$b/mq${X}$b$V.yaml > $OUT/c$b.log 2>&1) || { echo "compile$b failed, see $OUT/c$b.log"; exit 1; }
+done
+(cd $T && $PY -m esphome upload $M/$V/32/mq${X}32$V.yaml --device $P32 > $OUT/u32.log 2>&1) & U32=$!
+(cd $T && $PY -m esphome upload $M/$V/66/mq${X}66$V.yaml --device $P66 > $OUT/u66.log 2>&1) & U66=$!
+wait $U32 || { echo "upload32 failed, see $OUT/u32.log"; exit 1; }
+wait $U66 || { echo "upload66 failed, see $OUT/u66.log"; exit 1; }
+mosquitto_sub -h 127.0.0.1 -v -t '#' -F '@H:@M:@S r=%r %t %p' > $OUT/broker.txt 2>&1 &
+SUB=$!
+$PY $SC $P32 170 $OUT/s32.txt --reset > /dev/null 2>&1 &
+$PY $SC $P66 170 $OUT/s66.txt > /dev/null 2>&1 &
+sleep 60
+TOPICS=$(grep "homeassistant/button/" $OUT/broker.txt | grep "\"name\":\"mq${X}" | grep -o '"cmd_t":"[^"]*"' | cut -d'"' -f4 | sort -u)
+ERR=0
+[ "$(echo "$TOPICS" | grep -c .)" -ge 2 ] || { echo "ERROR: expected a Restart command topic per board, found: ${TOPICS:-none}"; ERR=1; }
+for t in $TOPICS; do
+  mosquitto_pub -h 127.0.0.1 -t "$t" -m PRESS && echo "$(date +%T) press $t" || { echo "$(date +%T) ERROR: publish to $t failed"; ERR=1; }
+done
+sleep 45
+pkill -f "[s]erial_capture.py $P32"; pkill -f "[s]erial_capture.py $P66"; sleep 1
+echo "$(date +%T) hold both in the bootloader for 40 s (will test)"
+$PY -m esptool --port $P32 --after no-reset chip-id > $OUT/hold32.log 2>&1 & H32=$!
+$PY -m esptool --port $P66 --after no-reset chip-id > $OUT/hold66.log 2>&1 & H66=$!
+wait $H32 || { echo "ERROR: ESP32 not held in the bootloader, no will expected (see $OUT/hold32.log)"; ERR=1; }
+wait $H66 || { echo "ERROR: ESP8266 not held in the bootloader, no will expected (see $OUT/hold66.log)"; ERR=1; }
+sleep 40
+REL=$(date +%T); echo "$REL release both"
+$PY $SC $P32 40 $OUT/s32b.txt --reset > /dev/null 2>&1 &
+$PY $SC $P66 40 $OUT/s66b.txt --reset > /dev/null 2>&1 &
+sleep 42
+kill $SUB
+for f in s32 s66; do echo "== $f"; grep -a -E "Topic Prefix|Log Topic|Availability|Client ID" $OUT/$f.txt | head -4 | sed 's/^\[[0-9:]*\]//'; done
+echo "== broker (excluding discovery and entity state)"
+grep -v -E " homeassistant/| [^ ]*/(sensor|button|binary_sensor)/| esphome/discover/" $OUT/broker.txt | awk '$3 ~ /(debug|log)$/ { if (seen[$3]++) next; print $1, $2, $3, "(log lines:", "see file)"; next } { print }' | head -40
+echo "== log line counts per topic"; awk '$3 ~ /(debug|log)$/ {c[$3]++} END {for (t in c) print t, c[t]}' $OUT/broker.txt
+# each board must publish its birth message (default "online", explicit variant "up") after the release;
+# a missing will is the measured result and does not fail the run
+BIRTHS=$(awk -v t="$REL" '$1 >= t && ($4 == "online" || $4 == "up") {print $3}' $OUT/broker.txt | sort -u | grep -c .)
+[ "$BIRTHS" -ge 2 ] || { echo "ERROR: $BIRTHS of 2 boards published a birth message after the release"; ERR=1; }
+exit $ERR

.agents/skills/hardware-test/scripts/mqtt/shutdown/base/sd32base.yaml

@@ -0,0 +1,9 @@
+substitutions: {name: sd32base}
+esphome:
+  name: sd32base
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+deep_sleep:
+  id: ds
+packages: {c: !include ../sd-common.yaml}

.agents/skills/hardware-test/scripts/mqtt/shutdown/base/sd66base.yaml

@@ -0,0 +1,8 @@
+substitutions: {name: sd66base}
+esphome:
+  name: sd66base
+esp8266:
+  board: d1_mini
+deep_sleep:
+  id: ds
+packages: {c: !include ../sd-common.yaml}

.agents/skills/hardware-test/scripts/mqtt/shutdown/match.py

@@ -0,0 +1,33 @@
+"""match.py OUTDIR: for each event, say whether a bye/dead for that board arrived within 5 s."""
+
+import pathlib
+import sys
+
+d = pathlib.Path(sys.argv[1])
+
+
+def secs(stamp: str) -> int:
+    h, m, s = (int(x) for x in stamp.split(":"))
+    return h * 3600 + m * 60 + s
+
+
+ev = [ln.split() for ln in (d / "events.txt").read_text(encoding="utf-8").splitlines()]
+br = [
+    f
+    for f in (
+        ln.split() for ln in (d / "broker.txt").read_text(encoding="utf-8").splitlines()
+    )
+    if len(f) >= 3 and f[2] in ("bye", "dead")
+]
+rows = {}
+for ts, kind, b in ev:
+    hit = [
+        p
+        for s, top, p in br
+        if top.startswith(f"sd{b}") and 0 <= secs(s) - secs(ts) <= 5
+    ]
+    rows.setdefault((b, kind), []).append(
+        "bye" if "bye" in hit else (hit[0] if hit else "-")
+    )
+for (b, kind), r in sorted(rows.items()):
+    print(f"{b} {kind:8s} {r.count('bye')}/{len(r)} bye   {' '.join(r)}")

.agents/skills/hardware-test/scripts/mqtt/shutdown/pr/sd32pr.yaml

@@ -0,0 +1,9 @@
+substitutions: {name: sd32pr}
+esphome:
+  name: sd32pr
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+deep_sleep:
+  id: ds
+packages: {c: !include ../sd-common.yaml}

.agents/skills/hardware-test/scripts/mqtt/shutdown/pr/sd66pr.yaml

@@ -0,0 +1,8 @@
+substitutions: {name: sd66pr}
+esphome:
+  name: sd66pr
+esp8266:
+  board: d1_mini
+deep_sleep:
+  id: ds
+packages: {c: !include ../sd-common.yaml}

.agents/skills/hardware-test/scripts/mqtt/shutdown/sd-common.yaml

@@ -0,0 +1,40 @@
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+logger:
+  level: DEBUG
+
+mqtt:
+  broker: !secret broker_ip
+  reboot_timeout: 0s
+  discovery: false
+  birth_message: {topic: "${name}/status", payload: hello}
+  will_message: {topic: "${name}/status", payload: dead}
+  shutdown_message: {topic: "${name}/status", payload: bye}
+  on_message:
+    - topic: ${name}/cmd
+      payload: restart
+      then:
+        - button.press: restart_btn
+    - topic: ${name}/cmd
+      payload: disable
+      then:
+        - mqtt.disable:
+        - delay: 5s
+        - mqtt.enable:
+    - topic: ${name}/cmd
+      payload: sleep
+      then:
+        - deep_sleep.enter:
+            id: ds
+            sleep_duration: 5s
+
+ota:
+  - platform: esphome
+
+button:
+  - platform: restart
+    id: restart_btn
+    name: Restart

.agents/skills/hardware-test/scripts/mqtt/shutdown/sdrun.sh

@@ -0,0 +1,44 @@
+#!/bin/bash
+# sdrun.sh V(pr|base) [N] : flash both boards, restart each N times over MQTT, then disable, then deep sleep (ESP32)
+# Copy this directory to a scratch directory (secrets.yaml with wifi_ssid, wifi_password and broker_ip next to
+# sd-common.yaml) and run it there with a local mosquitto on this host.
+# env: TREE (esphome checkout for V), P32 and P66 (ESP32 and ESP8266 serial ports), PY (checkout python)
+S=$(cd "$(dirname "$0")" && pwd); V=$1; N=${2:-10}; T=${TREE:?set TREE}; PY=${PY:-$T/venv/bin/python}
+P32=${P32:?set P32}; P66=${P66:?set P66}; OUT=$S/out-$V; rm -rf "$OUT"; mkdir -p "$OUT"
+pkill -f "[s]erial_capture.py $P32"; pkill -f "[s]erial_capture.py $P66"
+# compile one at a time (shared tool caches), then flash both boards at once; upload alone never builds
+for b in 32 66; do
+  (cd "$T" && $PY -m esphome compile "$S/$V/sd$b$V.yaml" > "$OUT/c$b.log" 2>&1) || { echo "compile$b failed, see $OUT/c$b.log"; exit 1; }
+done
+(cd "$T" && $PY -m esphome upload "$S/$V/sd32$V.yaml" --device "$P32" > "$OUT/u32.log" 2>&1) & U32=$!
+(cd "$T" && $PY -m esphome upload "$S/$V/sd66$V.yaml" --device "$P66" > "$OUT/u66.log" 2>&1) & U66=$!
+wait $U32 || { echo "upload32 failed"; exit 1; }; wait $U66 || { echo "upload66 failed"; exit 1; }
+mosquitto_sub -h 127.0.0.1 -R -v -t "sd32$V/status" -t "sd66$V/status" -F '@H:@M:@S %t %p' > "$OUT/broker.txt" 2>&1 & SUB=$!
+# cmd KIND BOARD: publish the command and record the event only when the publish worked
+ERR=0
+cmd() {
+  local ts; ts=$(date +%T)
+  if mosquitto_pub -h 127.0.0.1 -t "sd$2$V/cmd" -m "$1"; then echo "$ts $1 $2" >> "$OUT/events.txt"
+  else echo "$ts ERROR: publish $1 to board $2 failed"; ERR=1; fi
+}
+sleep 40
+for b in 32 66; do
+  grep -q "sd$b$V/status hello" "$OUT/broker.txt" || { echo "ERROR: sd$b$V never sent its birth message"; kill $SUB; exit 1; }
+done
+for i in $(seq "$N"); do
+  for b in 32 66; do cmd restart $b; done
+  sleep 25
+done
+for b in 32 66; do cmd disable $b; done
+sleep 30
+for i in 1 2 3; do cmd sleep 32; sleep 30; done
+kill -0 $SUB 2>/dev/null || { echo "ERROR: the broker subscriber died, counts are incomplete"; ERR=1; }
+kill $SUB 2>/dev/null
+for b in 32 66; do echo "sd$b$V: $(grep -c "sd$b$V/status hello" "$OUT/broker.txt") hello, $(grep -c "sd$b$V/status bye" "$OUT/broker.txt") bye, $(grep -c "sd$b$V/status dead" "$OUT/broker.txt") dead"; done
+# each restart, the disable (a reconnect) and each ESP32 deep sleep is followed by a new birth message
+for b in 32 66; do
+  want=$((N + 2)); [ $b = 32 ] && want=$((want + 3))
+  got=$(grep -c "sd$b$V/status hello" "$OUT/broker.txt")
+  [ "$got" -ge "$want" ] || { echo "ERROR: sd$b$V sent $got birth messages, expected $want (a board did not come back)"; ERR=1; }
+done
+exit $ERR

.agents/skills/hardware-test/scripts/online_image/firstpass.py

@@ -0,0 +1,115 @@
+"""Reboot the DUT before each fetch and report the longest online_image loop pass of that one fetch.
+usage: firstpass.py HOST KEY BASEURL LABEL IMAGE...
+The component warning prints only on a new maximum, so a fresh boot per image gives that image's own maximum.
+"""
+
+import asyncio
+import contextlib
+import re
+import sys
+import time
+
+from aioesphomeapi import (
+    APIClient,
+    ButtonInfo,
+    InvalidAuthAPIError,
+    InvalidEncryptionKeyAPIError,
+    LogLevel,
+    RequiresEncryptionAPIError,
+    UserService,
+)
+
+HOST, KEY, BASE, LABEL = sys.argv[1:5]
+IMAGES = sys.argv[5:]
+TOOK = re.compile(r"online_image\.image took a long time for an operation \((\d+) ms\)")
+OUTCOME = re.compile(r"download (finished|error): img\b")
+
+
+async def connect() -> tuple[APIClient, dict, int]:
+    last: Exception | None = None
+    for _ in range(30):
+        c = APIClient(HOST, 6053, None, noise_psk=KEY)
+        try:
+            await c.connect(login=True)
+            break
+        except (
+            InvalidAuthAPIError,
+            InvalidEncryptionKeyAPIError,
+            RequiresEncryptionAPIError,
+        ):
+            raise  # a wrong key never fixes itself
+        except Exception as exc:  # noqa: BLE001
+            last = exc
+            await asyncio.sleep(2)
+    else:
+        raise RuntimeError(f"device did not come back (last error {last!r})")
+    ents, services = await c.list_entities_services()
+    reboot = next(
+        (e.key for e in ents if isinstance(e, ButtonInfo) and e.name == "Reboot"), None
+    )
+    if reboot is None:
+        raise RuntimeError("the device has no button named Reboot")
+    return c, {s.name: s for s in services if isinstance(s, UserService)}, reboot
+
+
+async def rebooted(c: APIClient, reboot: int) -> bool:
+    """Press Reboot and wait up to 15 s for the API connection to drop."""
+    c.button_command(reboot)
+    end = time.monotonic() + 15
+    while time.monotonic() < end:
+        try:
+            await asyncio.wait_for(c.device_info(), 1.0)
+        except TimeoutError:
+            continue  # a slow reply is not a reboot; only a closed connection counts
+        except Exception:  # noqa: BLE001
+            return True
+        await asyncio.sleep(0.2)
+    return False
+
+
+async def main() -> int:
+    failed = []
+    for img in IMAGES:
+        c, actions, reboot = await connect()
+        ok = await rebooted(c, reboot)
+        with contextlib.suppress(Exception):
+            await c.disconnect()
+        if not ok:
+            print(f"{LABEL} {img:22} SKIPPED: no reboot within 15 s", flush=True)
+            failed.append(img)
+            continue
+        await asyncio.sleep(8)
+        c, actions, reboot = await connect()
+        lines: list[str] = []
+        c.subscribe_logs(
+            lambda m, lines=lines: lines.append(m.message.decode(errors="replace")),
+            log_level=LogLevel.LOG_LEVEL_DEBUG,
+        )
+        await asyncio.sleep(3)
+        lines.clear()
+        t0 = time.monotonic()
+        await c.execute_service(actions["fetch"], {"url": f"{BASE}/{img}"})
+        outcome = None
+        while time.monotonic() - t0 < 60 and outcome is None:
+            await asyncio.sleep(0.2)
+            outcome = next(
+                (m.group(1) for ln in lines if (m := OUTCOME.search(ln))), None
+            )
+        took = time.monotonic() - t0
+        await asyncio.sleep(1)
+        passes = [int(m.group(1)) for ln in lines if (m := TOOK.search(ln))]
+        print(
+            f"{LABEL} {img:22} {outcome or 'none':9} {took:5.1f}s longest pass {max(passes) if passes else '<50'} ms {passes}",
+            flush=True,
+        )
+        if (
+            outcome != "finished"
+        ):  # a decode timing is only meaningful for a complete image
+            failed.append(img)
+        await c.disconnect()
+    if failed:
+        print(f"{LABEL} not finished: {', '.join(failed)}", flush=True)
+    return 1 if failed else 0
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/online_image/make_big_pngs.py

@@ -0,0 +1,32 @@
+"""Build the large PNGs for the decode time test: a compressible gradient (also saved as ok.png so every server case
+carries a multi-row PNG), a flat 640x480 image whose few KB hold 307200 pixels, incompressible noise, and a JPEG."""
+
+from pathlib import Path
+import random
+import sys
+
+from PIL import Image
+
+www = Path(sys.argv[1])
+im = Image.new("RGB", (320, 240))
+px = im.load()
+for y in range(240):
+    for x in range(320):
+        px[x, y] = (x * 255 // 319, y * 255 // 239, 128)
+im.save(www / "ok.png", "PNG", optimize=True)
+im.save(www / "gradient.png", "PNG", optimize=True)
+im.save(www / "gradient.jpg", "JPEG", quality=85)
+Image.new("RGB", (640, 480), (40, 90, 200)).save(www / "flat.png", "PNG", optimize=True)
+random.seed(1)
+noise = Image.new("RGB", (320, 240))
+npx = noise.load()
+for y in range(240):
+    for x in range(320):
+        npx[x, y] = (
+            random.randrange(256),
+            random.randrange(256),
+            random.randrange(256),
+        )
+noise.save(www / "noise.png", "PNG")
+for p in sorted(www.iterdir()):
+    print(p.name, p.stat().st_size)

.agents/skills/hardware-test/scripts/online_image/make_bmp_qoi.py

@@ -0,0 +1,93 @@
+"""Write big BMP (1, 8, 24 bit) and QOI test images into a www directory."""
+
+from pathlib import Path
+import random
+import struct
+import sys
+
+from PIL import Image
+
+www = Path(sys.argv[1])
+
+
+def gradient(w, h):
+    im = Image.new("RGB", (w, h))
+    im.putdata(
+        [
+            ((x * 255) // w, (y * 255) // h, ((x + y) * 255) // (w + h))
+            for y in range(h)
+            for x in range(w)
+        ]
+    )
+    return im
+
+
+def checker(w, h):
+    im = Image.new("1", (w, h))
+    im.putdata(
+        [255 if ((x // 16) + (y // 16)) % 2 else 0 for y in range(h) for x in range(w)]
+    )
+    return im
+
+
+checker(640, 480).save(www / "bmp1_640.bmp")
+
+gradient(320, 240).convert("P", palette=Image.ADAPTIVE).save(www / "bmp8_320.bmp")
+gradient(320, 240).save(www / "bmp24_320.bmp")
+
+
+def qoi(im, path):
+    im = im.convert("RGBA")
+    w, h = im.size
+    out = bytearray(b"qoif" + struct.pack(">IIBB", w, h, 4, 0))
+    idx = [(0, 0, 0, 0)] * 64
+    prev = (0, 0, 0, 255)
+    run = 0
+    px = list(im.getdata())
+    for i, p in enumerate(px):
+        if p == prev:
+            run += 1
+            if run == 62 or i == len(px) - 1:
+                out.append(0xC0 | (run - 1))
+                run = 0
+            continue
+        if run:
+            out.append(0xC0 | (run - 1))
+            run = 0
+        h_ = (p[0] * 3 + p[1] * 5 + p[2] * 7 + p[3] * 11) % 64
+        if idx[h_] == p:
+            out.append(h_)
+        else:
+            idx[h_] = p
+            if p[3] == prev[3]:
+                dr, dg, db = (
+                    ((p[0] - prev[0] + 128) % 256) - 128,
+                    ((p[1] - prev[1] + 128) % 256) - 128,
+                    ((p[2] - prev[2] + 128) % 256) - 128,
+                )
+                drg, dbg = dr - dg, db - dg
+                if -2 <= dr <= 1 and -2 <= dg <= 1 and -2 <= db <= 1:
+                    out.append(0x40 | (dr + 2) << 4 | (dg + 2) << 2 | (db + 2))
+                elif -32 <= dg <= 31 and -8 <= drg <= 7 and -8 <= dbg <= 7:
+                    out += bytes([0x80 | (dg + 32), (drg + 8) << 4 | (dbg + 8)])
+                else:
+                    out += bytes([0xFE, p[0], p[1], p[2]])
+            else:
+                out += bytes([0xFF, *p])
+        prev = p
+    out += b"\x00" * 7 + b"\x01"
+    path.write_bytes(out)
+
+
+qoi(
+    Image.new("RGB", (640, 480), (0, 0, 0)).copy(), www / "qoi_flat.qoi"
+)  # all runs, 62 pixels a byte
+qoi(checker(640, 480).convert("RGB"), www / "qoi_check.qoi")
+qoi(gradient(320, 240), www / "qoi_grad.qoi")
+random.seed(1)
+noise = Image.frombytes(
+    "RGB", (320, 240), bytes(random.getrandbits(8) for _ in range(320 * 240 * 3))
+)
+qoi(noise, www / "qoi_noise.qoi")
+for f in sorted(www.glob("*_*.*")):
+    print(f.name, f.stat().st_size)

.agents/skills/hardware-test/scripts/online_image/make_jpegs.py

@@ -0,0 +1,42 @@
+"""Build the JPEGs for the streaming decode test: noise 320x240 (also ok.jpg) and 640x480, a 640x480 gradient, the
+320x240 noise with a 12 KB EXIF segment in front of the frame header, and a progressive one."""
+
+from pathlib import Path
+import random
+import sys
+
+from PIL import Image
+
+www = Path(sys.argv[1])
+random.seed(2)
+
+
+def noise(w: int, h: int) -> Image.Image:
+    im = Image.new("RGB", (w, h))
+    px = im.load()
+    for y in range(h):
+        for x in range(w):
+            px[x, y] = (
+                random.randrange(256),
+                random.randrange(256),
+                random.randrange(256),
+            )
+    return im
+
+
+n320 = noise(320, 240)
+n320.save(www / "noise320.jpg", "JPEG", quality=90)
+n320.save(www / "ok.jpg", "JPEG", quality=90)
+noise(640, 480).save(www / "noise640.jpg", "JPEG", quality=85)
+g = Image.new("RGB", (640, 480))
+px = g.load()
+for y in range(480):
+    for x in range(640):
+        px[x, y] = (x * 255 // 639, y * 255 // 479, 128)
+g.save(www / "gradient640.jpg", "JPEG", quality=85)
+n320.save(
+    www / "bigexif320.jpg", "JPEG", quality=90, exif=b"Exif\x00\x00" + bytes(12000)
+)
+n320.save(www / "progressive320.jpg", "JPEG", quality=90, progressive=True)
+for p in sorted(www.iterdir()):
+    print(p.name, p.stat().st_size)

.agents/skills/hardware-test/scripts/online_image/oi_server.py

@@ -5,7 +5,16 @@ from pathlib import Path
 import sys
 import time
 
-WWW = Path(sys.argv[1])
+WWW = Path(sys.argv[1]).resolve()
+NOT_FOUND = b"HTTP/1.1 404 Not Found\r\nContent-Length: 0\r\nConnection: close\r\n\r\n"
+
+
+def www_file(rel: str) -> Path | None:
+    """The file under WWW named by a request path, or None if missing or outside WWW."""
+    file = (WWW / rel).resolve()
+    return file if file.is_relative_to(WWW) and file.is_file() else None
+
+
 PORT = int(sys.argv[2])
 LOG = Path(sys.argv[3]).open("a", encoding="utf-8", buffering=1)  # noqa: SIM115
 
@@ -108,8 +117,34 @@ async def respond(path: str, req: bytes, w: asyncio.StreamWriter) -> str:
     if name == "huge":  # claims 20 MB so the JPEG buffer resize fails
         w.write(headers(ctype, [b"Content-Length: 20000000"]) + body)
         return f"claims 20000000, sent {len(body)} bytes"
-    w.write(b"HTTP/1.1 404 Not Found\r\nContent-Length: 0\r\nConnection: close\r\n\r\n")
+    prefix, _, rest = name.partition("_")
-    return "404"
+    if prefix in ("half", "noclen", "chunked", "stall") and rest:
+        # 404 when the base file is missing, so a typo never passes as the misbehaviour under test
+        file = www_file(rest + "." + ext)
+    else:
+        prefix, file = "", www_file(path.lstrip("/"))
+    if file is None:
+        w.write(NOT_FOUND)
+        return "404"
+    data = file.read_bytes()
+    if prefix == "half":  # the first half of a file, Content-Length matching
+        cut = data[: len(data) // 2]
+        w.write(headers(ctype, [b"Content-Length: %d" % len(cut)]) + cut)
+        return f"half {len(cut)} of {len(data)} bytes"
+    if prefix == "noclen":  # a file without Content-Length, delimited by close
+        w.write(headers(ctype, []) + data)
+        return f"no content-length, full {len(data)} bytes"
+    if prefix == "chunked":  # a file in 1 KB chunks
+        w.write(headers(ctype, [b"Transfer-Encoding: chunked"]) + chunked(data, 1024))
+        return f"chunked full {len(data)} bytes"
+    if prefix == "stall":  # claims the full length, sends a third, holds the socket
+        part = data[: len(data) // 3]
+        w.write(headers(ctype, [b"Content-Length: %d" % len(data)]) + part)
+        await w.drain()
+        await asyncio.sleep(40)
+        return f"stalled after {len(part)} of {len(data)} bytes"
+    w.write(headers(ctype, [b"Content-Length: %d" % len(data), *tag]) + data)
+    return f"file {file.name} {len(data)} bytes"
 
 
 async def handle(r: asyncio.StreamReader, w: asyncio.StreamWriter) -> None:

.agents/skills/hardware-test/scripts/online_image/pngdrive.py

@@ -0,0 +1,261 @@
+"""Fetch images through the online_image API actions and report outcome, time, loop time maximum and heap minimum per case.
+usage: [CASES=png|noise|end|bmpqoi|short|jpeg] pngdrive.py HOST KEY LOGFILE BASEURL LABEL
+"""
+
+import asyncio
+import contextlib
+import os
+import pathlib
+import re
+import sys
+import time
+import urllib.error
+import urllib.request
+
+from aioesphomeapi import APIClient, LogLevel, SensorInfo, SensorState, UserService
+
+HOST, KEY, LOGFILE, BASE, LABEL = sys.argv[1:6]
+CAP_S = 60
+OUTCOME = re.compile(r"download (finished|error): (\w+)")
+INTEREST = re.compile(
+    r"took a long time|Error|error|Download ended|No data received|Waiting for more|Decoded|cut off|already being|headers do not fit|buffer_size|Downloading image|fully downloaded|allocat|not caching"
+)
+
+PNG = [
+    ("gradient 320x240 png", "fetch", "gradient.png", "finished"),
+    ("gradient again", "fetch", "gradient.png", "finished"),
+    ("gradient again", "fetch", "gradient.png", "finished"),
+    ("flat 640x480 png", "fetch", "flat.png", "finished"),
+    ("flat again", "fetch", "flat.png", "finished"),
+    ("flat again", "fetch", "flat.png", "finished"),
+    ("gradient jpeg", "fetch", "gradient.jpg", "finished"),
+    ("ok bmp 16x16", "fetch", "ok.bmp", "finished"),
+    ("ok png 16x16 body = gradient", "fetch", "ok.png", "finished"),
+    ("chunked png", "fetch", "chunked.png", "finished"),
+    ("chunked png cut in half", "fetch", "chunked_cut.png", "any"),
+    ("closed early png", "fetch", "closed_early.png", "error"),
+    ("stalled png", "fetch", "stall.png", "error"),
+    ("slow png", "fetch", "slow.png", "finished"),
+    ("empty png", "fetch", "empty.png", "error"),
+    ("no content-length png", "fetch", "noclen.png", "any"),
+    ("truncated bmp", "fetch", "truncated.bmp", "error"),
+    ("gradient after errors", "fetch", "gradient.png", "finished"),
+    ("gradient refetch 304", "refetch", "", "finished"),
+    ("noise 320x240 png", "fetch", "noise.png", "finished"),
+    ("noise again", "fetch", "noise.png", "finished"),
+]
+
+NOISE = [("noise 320x240 png", "fetch", "noise.png", "finished")] * 3
+
+END = [
+    (f"{pre}{img}", "fetch", f"{pre}{img}", exp)
+    for img in ("flat.png", "gradient.png")
+    for pre, exp in (
+        ("", "finished"),
+        ("noclen_", "finished"),
+        ("chunked_", "finished"),
+        ("half_", "any"),
+        ("stall_", "error"),
+    )
+] + [("gradient after", "fetch", "gradient.png", "finished")]
+
+BMPQOI = [
+    (f"{pre}{img}", "fetch", f"{pre}{img}", exp)
+    for img in (
+        "bmp1_640.bmp",
+        "bmp8_320.bmp",
+        "bmp24_320.bmp",
+        "qoi_flat.qoi",
+        "qoi_check.qoi",
+        "qoi_grad.qoi",
+        "qoi_noise.qoi",
+    )
+    for pre, exp in (
+        ("", "finished"),
+        ("", "finished"),
+        ("noclen_", "any"),
+        ("chunked_", "any"),
+        ("half_", "error"),
+    )
+] + [
+    ("ok bmp 16x16", "fetch", "ok.bmp", "finished"),
+    ("ok qoi 16x16", "fetch", "ok.qoi", "finished"),
+    ("truncated bmp", "fetch", "truncated.bmp", "error"),
+    ("stall_bmp1_640.bmp", "fetch", "stall_bmp1_640.bmp", "error"),
+    ("stall_qoi_flat.qoi", "fetch", "stall_qoi_flat.qoi", "error"),
+    ("qoi_flat after", "fetch", "qoi_flat.qoi", "finished"),
+]
+
+SHORT = [
+    ("gradient 320x240 png", "fetch", "gradient.png", "finished"),
+    ("gradient again", "fetch", "gradient.png", "finished"),
+    ("noise 320x240 png", "fetch", "noise.png", "finished"),
+    ("noise again", "fetch", "noise.png", "finished"),
+    ("noise again", "fetch", "noise.png", "finished"),
+    ("gradient after noise", "fetch", "gradient.png", "finished"),
+]
+
+JPEG = [
+    ("noise 320x240 jpeg 69 KB", "fetch", "noise320.jpg", "finished"),
+    ("noise 320x240 again", "fetch", "noise320.jpg", "finished"),
+    ("noise 320x240 again", "fetch", "noise320.jpg", "finished"),
+    ("noise 640x480 jpeg 233 KB", "fetch", "noise640.jpg", "finished"),
+    ("noise 640x480 again", "fetch", "noise640.jpg", "finished"),
+    ("gradient 640x480 jpeg 14 KB", "fetch", "gradient640.jpg", "finished"),
+    ("4 KB buffer: noise 320x240", "fetch_small", "noise320.jpg", "finished"),
+    ("4 KB buffer: noise 640x480", "fetch_small", "noise640.jpg", "finished"),
+    ("4 KB buffer: 12 KB exif", "fetch_small", "bigexif320.jpg", "any"),
+    ("64 KB buffer: 12 KB exif", "fetch", "bigexif320.jpg", "finished"),
+    ("progressive jpeg", "fetch", "progressive320.jpg", "any"),
+    ("no content-length", "fetch", "noclen_noise320.jpg", "finished"),
+    ("chunked", "fetch", "chunked_noise320.jpg", "finished"),
+    ("half with matching length", "fetch", "half_noise320.jpg", "error"),
+    ("closed early", "fetch", "closed_early.jpg", "error"),
+    ("stalled after a third", "fetch", "stall_noise320.jpg", "error"),
+    ("empty", "fetch", "empty.jpg", "error"),
+    ("noise 320x240 after errors", "fetch", "noise320.jpg", "finished"),
+    ("refetch 304", "refetch", "", "finished"),
+    ("ok png 16x16", "fetch", "ok.png", "finished"),
+    ("ok bmp 16x16", "fetch", "ok.bmp", "finished"),
+    ("4 KB buffer: noise 640x480 again", "fetch_small", "noise640.jpg", "finished"),
+    ("4 KB buffer: half", "fetch_small", "half_noise320.jpg", "error"),
+    ("4 KB buffer: stalled", "fetch_small", "stall_noise320.jpg", "error"),
+    (
+        "4 KB buffer: noise 320x240 after errors",
+        "fetch_small",
+        "noise320.jpg",
+        "finished",
+    ),
+]
+
+CASE_SETS = {
+    "png": PNG,
+    "noise": NOISE,
+    "end": END,
+    "bmpqoi": BMPQOI,
+    "short": SHORT,
+    "jpeg": JPEG,
+}
+CASES = CASE_SETS[os.environ.get("CASES", "png")]
+
+
+def missing_base_files() -> list[str]:
+    """Base files behind prefixed cases that the server does not serve (any HTTP error, or no server).
+
+    A failed fetch ends as a download error on the device, which would pass the half_ and stall_ cases untested.
+    """
+    builtin = {"chunked_cut", "closed_early", "large_header"}
+    missing = []
+    for _, _, path, _ in CASES:
+        stem = path.rsplit(".", 1)[0]
+        prefix, _, rest = path.partition("_")
+        if (
+            stem in builtin
+            or prefix not in ("half", "noclen", "chunked", "stall")
+            or not rest
+        ):
+            continue
+        try:
+            with urllib.request.urlopen(f"{BASE}/{rest}", timeout=10):
+                pass
+        except (urllib.error.URLError, OSError) as exc:  # HTTPError is a URLError
+            missing.append(f"{rest} ({exc})")
+    return sorted(set(missing))
+
+
+async def main() -> int:
+    if missing := missing_base_files():
+        print(
+            f"server has no {', '.join(missing)}; run the image generators first",
+            flush=True,
+        )
+        return 2
+    c = APIClient(HOST, 6053, None, noise_psk=KEY)
+    await c.connect(login=True)
+    ents, services = await c.list_entities_services()
+    actions = {s.name: s for s in services if isinstance(s, UserService)}
+    loop_key = next(
+        e.key for e in ents if isinstance(e, SensorInfo) and e.name == "Loop Time"
+    )
+    heap_key = next(
+        e.key for e in ents if isinstance(e, SensorInfo) and e.name == "Heap Free"
+    )
+    state = {"loop": [], "heap": None, "heap_min": None, "lines": []}
+
+    def on_state(s) -> None:
+        if isinstance(s, SensorState) and s.key == loop_key:
+            state["loop"].append(s.state)
+        elif isinstance(s, SensorState) and s.key == heap_key:
+            state["heap"] = s.state
+            if state["heap_min"] is None or s.state < state["heap_min"]:
+                state["heap_min"] = s.state
+
+    def on_log(msg) -> None:
+        line = msg.message.decode(errors="replace")
+        log.write(f"{time.strftime('%H:%M:%S')} {line}\n")
+        state["lines"].append(line)
+
+    with pathlib.Path(LOGFILE).open("w", encoding="utf-8") as log:
+        c.subscribe_states(on_state)
+        c.subscribe_logs(on_log, log_level=LogLevel.LOG_LEVEL_DEBUG)
+        await asyncio.sleep(3)
+        print(f"{LABEL}: heap at start {state['heap']}")
+        rows = []
+        for label, action, path, expect in CASES:
+            img = "img_small" if action.endswith("_small") else "img"
+            state["loop"].clear()
+            state["lines"].clear()
+            state["heap_min"] = None
+            args = {} if action.startswith("refetch") else {"url": f"{BASE}/{path}"}
+            t0 = time.monotonic()
+            log.write(f">>> {label} {path}\n")
+            try:
+                await c.execute_service(actions[action], args)
+            except Exception as exc:  # noqa: BLE001
+                print(f"      api call failed ({exc!r}); reconnecting", flush=True)
+                with contextlib.suppress(Exception):
+                    await c.disconnect()
+                await asyncio.sleep(10)
+                await c.connect(login=True)
+                ents, services = await c.list_entities_services()
+                actions = {s.name: s for s in services if isinstance(s, UserService)}
+                c.subscribe_states(on_state)
+                c.subscribe_logs(on_log, log_level=LogLevel.LOG_LEVEL_DEBUG)
+                await asyncio.sleep(2)
+                state["lines"].clear()
+                t0 = time.monotonic()
+                await c.execute_service(actions[action], args)
+            outcome = "no outcome"
+            while time.monotonic() - t0 < CAP_S:
+                await asyncio.sleep(0.25)
+                for line in state["lines"]:
+                    m = OUTCOME.search(line)
+                    if m and m.group(2) == img:
+                        outcome = m.group(1)
+                if outcome != "no outcome":
+                    break
+            took = time.monotonic() - t0
+            await asyncio.sleep(4.5)  # two loop time reports after the case
+            notes = [
+                ln for ln in state["lines"] if INTEREST.search(ln) and "Loop" not in ln
+            ]
+            loop_max = max(state["loop"]) if state["loop"] else float("nan")
+            ok = outcome == expect or (expect == "any" and outcome != "no outcome")
+            rows.append((label, path, expect, outcome, took, loop_max, ok))
+            print(
+                f"{'PASS' if ok else 'FAIL'} {label:30s} {path:18s} expected {expect:9s} got {outcome:10s} {took:5.1f}s loop max {loop_max:7.0f} ms heap min {state['heap_min']}",
+                flush=True,
+            )
+            for n in notes[:8]:
+                print(f"      {n.strip()}", flush=True)
+        await c.disconnect()
+    print(
+        "\nSUMMARY",
+        "all expectations met"
+        if all(r[6] for r in rows)
+        else "MISMATCHES: " + ", ".join(r[0] for r in rows if not r[6]),
+    )
+    return 0 if all(r[6] for r in rows) else 1
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/remote_base/rbdrive.py

@@ -0,0 +1,75 @@
+"""Run the tx_all script N times over the API and count what each sent frame decoded as.
+Usage: rbdrive.py IP N"""
+
+import asyncio
+import collections
+import re
+import sys
+
+import aioesphomeapi
+
+IP, N = sys.argv[1], int(sys.argv[2])
+lines: list[str] = []
+
+
+async def main() -> None:
+    c = aioesphomeapi.APIClient(IP, 6053, None)
+    await c.connect(login=True)
+    i = await c.device_info()
+    print("device", i.name, i.compilation_time, flush=True)
+    _, svcs = await c.list_entities_services()
+    s = {x.name: x for x in svcs}
+
+    def on_log(m):
+        lines.append(re.sub(r"\x1b\[[0-9;]*m", "", m.message.decode(errors="replace")))
+
+    c.subscribe_logs(on_log, log_level=aioesphomeapi.LogLevel.LOG_LEVEL_DEBUG)
+    await asyncio.sleep(1)
+    complete = 0
+    for rnd in range(1, N + 1):
+        start = len(lines)
+        await c.execute_service(s["tx_all"], {})
+        for _ in range(200):
+            if any("tx done" in x for x in lines[start:]):
+                complete += 1
+                break
+            await asyncio.sleep(0.1)
+        else:
+            print(f"round {rnd}: no tx done within 20 s", flush=True)
+        await asyncio.sleep(1)
+    await c.disconnect()
+    # per sent protocol: what was decoded
+    table: dict[str, collections.Counter] = collections.defaultdict(collections.Counter)
+    cur = None
+    for x in lines:
+        if m := re.search(r"\]: tx (\w+)$", x):
+            cur = m[1]
+            continue
+        if cur is None or cur == "done":
+            continue
+        if m := re.search(r"Received ([A-Za-z0-9_ ]+?):", x):
+            table[cur][m[1]] += 1
+        elif "ext dumper saw" in x:
+            table[cur]["ext"] += 1
+    print(f"{complete} of {N} rounds complete", flush=True)
+    for proto, cnt in table.items():
+        print(
+            f"{proto:11s} " + ", ".join(f"{k}={v}" for k, v in sorted(cnt.items())),
+            flush=True,
+        )
+    for x in lines:
+        if "ext dumper saw" not in x and re.search(
+            r"rbt|error|Error|overflow|Guru|panic", x
+        ):
+            print("  ", x[:160])
+
+    if not any(k != "ext" for cnt in table.values() for k in cnt):
+        print("nothing decoded", flush=True)
+        return 1
+    if not any(cnt["ext"] for cnt in table.values()):
+        print("the external dumper saw no frame", flush=True)
+        return 1
+    return 0 if complete == N else 1
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/remote_base/run.sh

@@ -0,0 +1,24 @@
+#!/bin/bash
+# run.sh LABEL B(32|66) [N] : write rb<B><LABEL> from templates/remote_base on every run, flash it, run rbdrive.py for N rounds
+# Copy this directory and templates/remote_base into one scratch directory (secrets.yaml next to rb-common.yaml).
+# env: TREE (esphome checkout to build from), P32 or P66 (serial port of the board), SC (path to
+# serial_capture.py), PY (checkout python), EXPECT_REGISTER=1 (fail when the dumper registration was rejected)
+R=$(cd "$(dirname "$0")" && pwd); V=$1; B=$2; N=${3:-3}; NAME=rb$B$V; T=${TREE:?set TREE}
+PY=${PY:-$T/venv/bin/python}; SC=${SC:?set SC}
+if [ "$B" = 32 ]; then P=${P32:?set P32}; else P=${P66:?set P66}; fi
+mkdir -p "$R/$V"; sed "s/@V@/$V/g" "$R/rb$B.tmpl" > "$R/$V/$NAME.yaml"  # always from the current template
+OUT=$R/out-$V-$B; rm -rf "$OUT"; mkdir -p "$OUT"
+pkill -f "[s]erial_capture.py $P"
+(cd "$T" && $PY -m esphome run --no-logs "$R/$V/$NAME.yaml" --device "$P" > "$OUT/upload.log" 2>&1) || { echo "build or upload failed"; exit 1; }
+$PY "$SC" "$P" 150 "$OUT/serial.txt" --reset > /dev/null 2>&1 & CAP=$!
+trap 'kill $CAP 2>/dev/null' EXIT
+for i in $(seq 60); do IP=$(grep -a -o 'IP Address: [0-9.]*' "$OUT/serial.txt" 2>/dev/null | tail -1 | cut -d' ' -f3); [ -z "$IP" ] && IP=$(python3 -c "import socket;print(socket.gethostbyname('$NAME.local'))" 2>/dev/null); [ -n "$IP" ] && break; sleep 2; done
+[ -n "$IP" ] || { echo "device not found"; exit 1; }
+$PY "$R/rbdrive.py" "$IP" "$N" > "$OUT/drive.txt" 2>&1; RC=$?
+pkill -f "[s]erial_capture.py $P"
+[ -s "$OUT/serial.txt" ] || { echo "serial capture got nothing, crash check invalid"; exit 1; }
+CRASH=$(grep -a -c -i 'guru\|panic\|exception\|rst cause' "$OUT/serial.txt")
+REJECTED=$(grep -a -c 'No dumper slot' "$OUT/serial.txt")
+cat "$OUT/drive.txt"; echo "serial: $(grep -a -c 'ext dumper register called' "$OUT/serial.txt") register calls, $REJECTED rejected, $CRASH crash lines"
+if [ "${EXPECT_REGISTER:-0}" = 1 ] && [ "$REJECTED" != 0 ]; then echo "the dumper registration was rejected"; exit 1; fi
+[ "$RC" = 0 ] && [ "$CRASH" = 0 ]

.agents/skills/hardware-test/scripts/uart_sim/frdrive.py

@@ -0,0 +1,103 @@
+"""Drive the FrameReader simulation builds over the API: press buttons, set selects/numbers, print states.
+usage: frdrive.py HOST LOGFILE OP [OP ...]
+  OP: list | press:NAME | sel:NAME=OPTION | num:NAME=VALUE | wait:SECONDS
+"""
+
+import asyncio
+import math
+import pathlib
+import sys
+import time
+
+from aioesphomeapi import (
+    APIClient,
+    BinarySensorState,
+    ButtonInfo,
+    LogLevel,
+    NumberInfo,
+    SelectInfo,
+    SensorInfo,
+    SensorState,
+)
+
+HOST, LOGFILE = sys.argv[1], sys.argv[2]
+OPS = sys.argv[3:]
+T0 = time.time()
+SKIP = {"Heap Free", "Loop Time"}
+
+
+def stamp() -> str:
+    return f"{time.time() - T0:7.3f}"
+
+
+async def main() -> int:
+    with pathlib.Path(LOGFILE).open("w", encoding="utf-8") as log:
+        c = APIClient(HOST, 6053, None)
+        await c.connect(login=True)
+        info = await c.device_info()
+        print(f"device {info.name} compiled {info.compilation_time}", flush=True)
+        ents, _ = await c.list_entities_services()
+        by_key = {e.key: e for e in ents}
+
+        def find(kind: type, name: str):
+            for e in ents:
+                if isinstance(e, kind) and e.name == name:
+                    return e
+            sys.exit(f"no {kind.__name__} named {name!r}")
+
+        def on_state(s) -> None:
+            e = by_key.get(s.key)
+            if e is None or e.name in SKIP:
+                return
+            if isinstance(s, SensorState) and isinstance(e, SensorInfo):
+                v = "nan" if math.isnan(s.state) else f"{s.state:.2f}"
+                line = f"{stamp()} SEN {e.name} = {v}"
+            elif isinstance(s, BinarySensorState):
+                line = f"{stamp()} BIN {e.name} = {s.state}"
+            else:
+                return
+            print(line, flush=True)
+            log.write(line + "\n")
+            states.append(line)
+
+        def on_log(msg) -> None:
+            log.write(f"{stamp()} {msg.message.decode(errors='replace')}\n")
+            log.flush()
+
+        states: list[str] = []
+        c.subscribe_states(on_state)
+        c.subscribe_logs(on_log, log_level=LogLevel.LOG_LEVEL_VERBOSE)
+        await asyncio.sleep(2)
+        states.clear()  # count only states published while the ops ran
+        for op in OPS:
+            kind, _, rest = op.partition(":")
+            if kind == "list":
+                for e in ents:
+                    print(type(e).__name__, e.name)
+                continue
+            if kind == "wait":
+                await asyncio.sleep(float(rest))
+                continue
+            line = f"{stamp()} >>> {op}"
+            print(line, flush=True)
+            log.write(line + "\n")
+            if kind == "press":
+                c.button_command(find(ButtonInfo, rest).key)
+            elif kind == "sel":
+                name, _, value = rest.partition("=")
+                c.select_command(find(SelectInfo, name).key, value)
+            elif kind == "num":
+                name, _, value = rest.partition("=")
+                c.number_command(find(NumberInfo, name).key, float(value))
+            else:
+                sys.exit(f"unknown op {op!r}")
+            await asyncio.sleep(2.5)
+        await asyncio.sleep(1)
+        await c.disconnect()
+    if any(not op.startswith(("list", "wait")) for op in OPS) and not states:
+        print("no sensor state was published while the ops ran", flush=True)
+        return 1
+    return 0
+
+
+sys.exit(asyncio.run(main()))

.agents/skills/hardware-test/scripts/uart_sim/run.sh

@@ -0,0 +1,19 @@
+#!/bin/bash
+# usage: run.sh NAME TREE SECONDS OP...   flash over serial, capture serial, drive over the API
+# Copy this, frdrive.py and the templates/uart_sim files into one scratch directory and run it there.
+# env: PORT (the devkit serial port), SC (path to serial_capture.py), PY (checkout python); TREE is the esphome
+# checkout to build from
+D=$(cd "$(dirname "$0")" && pwd); N=$1; T=$2; S=$3; shift 3; P=${PORT:?set PORT}; PY=${PY:-$T/venv/bin/python}
+(cd $T && PYTHONUNBUFFERED=1 timeout 1500 $PY -m esphome run --no-logs $D/$N.yaml --device $P > $D/$N-upload.log 2>&1) || { echo "build or upload failed"; exit 1; }
+$PY ${SC:?set SC} $P $S $D/$N-cap.txt > $D/$N-cap.err 2>&1 &
+CAP=$!
+trap 'kill $CAP 2>/dev/null' EXIT
+UP=0
+for i in $(seq 1 40); do $PY -c "import socket;socket.create_connection(('$N.local',6053),2)" 2>/dev/null && { UP=1; break; }; sleep 2; done
+[ $UP = 1 ] || { echo "device $N.local not reachable on port 6053"; exit 1; }
+cd $D && PYTHONUNBUFFERED=1 timeout $((S - 5)) $PY frdrive.py $N.local $N-api.log "$@" > $N-drive.txt 2>&1; RC=$?; echo "drive rc=$RC"
+wait $CAP; CAPRC=$?; trap - EXIT
+[ "$CAPRC" = 0 ] || { echo "serial capture failed (rc $CAPRC), crash check invalid"; exit 1; }
+CRASH=$(grep -acE 'Guru|abort\(\)|rst:0x' $D/$N-cap.txt)
+echo "crash lines: $CRASH"
+[ "$RC" = 0 ] && [ "$CRASH" = 0 ]

.agents/skills/hardware-test/templates/ble_midi/midicli.yaml

@@ -0,0 +1,150 @@
+substitutions:
+  # the fake instrument's BLE address; pass it with `-s peri_mac AA:BB:CC:DD:EE:FF`
+  peri_mac: "00:00:00:00:00:00"
+
+esphome:
+  name: midicli
+
+esp32:
+  board: esp32dev
+  framework:
+    type: esp-idf
+
+logger:
+  level: VERBOSE
+  logs:
+    esp32_ble_tracker: DEBUG
+    ble_client: DEBUG
+    esp32_ble: DEBUG
+    api: DEBUG
+    scheduler: DEBUG
+    sensor: DEBUG
+    wifi: DEBUG
+
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+api:
+  reboot_timeout: 0s
+
+esp32_ble_tracker:
+
+ble_client:
+  - mac_address: ${peri_mac}
+    id: midi_client
+
+ble_midi:
+  id: test_midi
+  ble_client_id: midi_client
+  max_sysex_size: 64
+  pair: false
+  on_connect:
+    - logger.log: "MIDI device connected"
+    - ble_midi.send_program_change:
+        id: test_midi
+        program: 0
+  on_disconnect:
+    - logger.log: "MIDI device disconnected"
+  on_note_on:
+    - logger.log:
+        format: "Note on %u velocity %u channel %u"
+        args: [note, velocity, channel]
+  on_note_off:
+    - logger.log:
+        format: "Note off %u channel %u"
+        args: [note, channel]
+  on_control_change:
+    - logger.log:
+        format: "CC %u = %u on channel %u"
+        args: [control_number, value, channel]
+  on_program_change:
+    - logger.log:
+        format: "Program %u on channel %u"
+        args: [program, channel]
+  on_pitch_bend:
+    - logger.log:
+        format: "Pitch bend %d on channel %u"
+        args: [value, channel]
+  on_sysex:
+    - lambda: |-
+        ESP_LOGI("midi", "SysEx of %u bytes: %s", (unsigned) data.size(), format_hex_pretty(data.data(), data.size()).c_str());
+  on_message:
+    - lambda: |-
+        ESP_LOGI("midi", "Raw message of %u bytes: %s", (unsigned) data.size(), format_hex_pretty(data.data(), data.size()).c_str());
+
+sensor:
+  - platform: debug
+    free:
+      name: Heap Free
+    loop_time:
+      name: Loop Time
+
+debug:
+  update_interval: 30s
+
+button:
+  - platform: template
+    name: "MIDI send tests"
+    on_press:
+      - ble_midi.send_note_on:
+          id: test_midi
+          note: 60
+      - ble_midi.send_note_on:
+          id: test_midi
+          note: !lambda "return 60;"
+          velocity: !lambda "return 100;"
+          channel: !lambda "return 1;"
+      - ble_midi.send_note_off:
+          id: test_midi
+          note: 60
+          velocity: 0
+          channel: 1
+      - ble_midi.send_control_change:
+          id: test_midi
+          control_number: 7
+          value: 64
+      - ble_midi.send_pitch_bend:
+          id: test_midi
+          value: -8192
+      - ble_midi.send_sysex:
+          id: test_midi
+          payload: [0x7E, 0x7F, 0x06, 0x01]
+      - ble_midi.send_raw:
+          id: test_midi
+          data: !lambda "return std::vector<uint8_t>{0x90, 0x3C, 0x40};"
+      - if:
+          condition:
+            ble_midi.connected:
+              id: test_midi
+          then:
+            - logger.log: "Still connected"
+          else:
+            - logger.log: "Not connected"
+  - platform: template
+    name: "MIDI send out of range"
+    on_press:
+      - ble_midi.send_note_on:
+          id: test_midi
+          note: !lambda "return 200;"
+      - ble_midi.send_control_change:
+          id: test_midi
+          control_number: 1
+          value: 1
+          channel: !lambda "return 16;"
+      - ble_midi.send_pitch_bend:
+          id: test_midi
+          value: 8191
+      - ble_midi.send_sysex:
+          id: test_midi
+          payload: [0xF0, 0x7E, 0x7F, 0x09, 0x01, 0xF7]
+      - ble_midi.send_raw:
+          id: test_midi
+          data: !lambda "return std::vector<uint8_t>(140, 0x00);"
+  - platform: template
+    name: "MIDI send long sysex"
+    on_press:
+      - ble_midi.send_sysex:
+          id: test_midi
+          payload: !lambda "return std::vector<uint8_t>(100, 0x11);"

.agents/skills/hardware-test/templates/ble_midi/midiperi.yaml

@@ -0,0 +1,177 @@
+esphome:
+  name: midiperi
+
+esp32:
+  board: esp32-s3-devkitc-1
+  variant: esp32s3
+  flash_size: 16MB
+  framework:
+    type: esp-idf
+
+logger:
+  hardware_uart: USB_SERIAL_JTAG
+  level: DEBUG
+
+api:
+  reboot_timeout: 0s
+
+ota:
+  - platform: esphome
+
+ethernet:
+  type: W5500
+  clk_pin: GPIO13
+  mosi_pin: GPIO11
+  miso_pin: GPIO12
+  cs_pin: GPIO14
+  interrupt_pin: GPIO10
+  reset_pin: GPIO9
+
+esp32_ble_server:
+  id: ble_server
+  manufacturer: FakeMIDI
+  on_connect:
+    - lambda: ESP_LOGI("peri", "central connected %d", id);
+  on_disconnect:
+    - lambda: ESP_LOGI("peri", "central disconnected %d", id);
+  services:
+    - uuid: 03B80E5A-EDE8-4B33-A751-6CE34EC4C700
+      advertise: true
+      characteristics:
+        - id: midi_io
+          uuid: 7772E5DB-3868-4112-A1A9-F2669D106BF3
+          read: true
+          write: true
+          write_no_response: true
+          notify: true
+          on_write:
+            - lambda: |-
+                ESP_LOGI("peri", "RX %u bytes: %s", (unsigned) x.size(), format_hex_pretty(x.data(), x.size()).c_str());
+
+script:
+  - id: emit
+    parameters:
+      pkt: std::vector<uint8_t>
+    then:
+      - lambda: |-
+          ESP_LOGI("peri", "TX %u bytes: %s", (unsigned) pkt.size(), format_hex_pretty(pkt.data(), pkt.size()).c_str());
+          id(midi_io).set_value(std::vector<uint8_t>(pkt));
+          id(midi_io).notify();
+
+button:
+  - platform: restart
+    name: P Restart
+  - platform: template
+    name: P Note On
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0x90, 0x3C, 0x64};'
+  - platform: template
+    name: P Note Off
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0x80, 0x3C, 0x00};'
+  - platform: template
+    name: P Control Change
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0xB0, 0x07, 0x40};'
+  - platform: template
+    name: P Program Change
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0xC5, 0x05};'
+  - platform: template
+    name: P Pitch Bend
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0xE0, 0x7F, 0x7F};'
+  - platform: template
+    name: P Two Notes One Packet
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0x90, 0x3C, 0x64, 0x81, 0x91, 0x3E, 0x50};'
+  - platform: template
+    name: P Running Status
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0x90, 0x40, 0x64, 0x41, 0x64};'
+  - platform: template
+    name: P Sysex One Packet
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0xF0, 0x7E, 0x7F, 0x06, 0x01, 0x80, 0xF7};'
+  - platform: template
+    name: P Sysex Split
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0xF0, 0x41, 0x10, 0x42};'
+      - delay: 200ms
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x12, 0x00, 0x40, 0x80, 0xF7};'
+  - platform: template
+    name: P Sysex Oversize
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'std::vector<uint8_t> v{0x80, 0x80, 0xF0}; for (int i = 0; i < 17; i++) v.push_back(i); return v;'
+      - delay: 200ms
+      - script.execute:
+          id: emit
+          pkt: !lambda 'std::vector<uint8_t> v{0x80}; for (int i = 0; i < 19; i++) v.push_back(0x20 + i); return v;'
+      - delay: 200ms
+      - script.execute:
+          id: emit
+          pkt: !lambda 'std::vector<uint8_t> v{0x80}; for (int i = 0; i < 19; i++) v.push_back(0x40 + i); return v;'
+      - delay: 200ms
+      - script.execute:
+          id: emit
+          pkt: !lambda 'std::vector<uint8_t> v{0x80}; for (int i = 0; i < 19; i++) v.push_back(0x60 + i); return v;'
+      - delay: 200ms
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0xF7};'
+  - platform: template
+    name: P Sysex Cut Short
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0xF0, 0x41, 0x10, 0x80, 0x90, 0x3C, 0x64};'
+  - platform: template
+    name: P Clock
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0xF8};'
+  - platform: template
+    name: P Aftertouch
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0xD3, 0x45};'
+  - platform: template
+    name: P Orphan Data Byte
+    on_press:
+      - script.execute:
+          id: emit
+          pkt: !lambda 'return {0x80, 0x80, 0x3C, 0x64};'
+  - platform: template
+    name: P Burst
+    on_press:
+      - repeat:
+          count: 50
+          then:
+            - script.execute:
+                id: emit
+                pkt: !lambda 'return {0x80, 0x80, 0x90, 0x30, 0x40, 0x80, 0x80, 0x30, 0x00};'
+            - delay: 20ms

.agents/skills/hardware-test/templates/http_request_action/hr-common.yaml

@@ -0,0 +1,162 @@
+# Package for the http_request action test (esphome#21486). Needs substitutions: name, bufsize.
+esphome:
+  name: ${name}
+
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+api:
+  reboot_timeout: 0s
+  actions:
+    - action: req_get
+      variables:
+        url: string
+      then:
+        - lambda: 'id(max_gap) = 0; id(t_start) = millis(); id(tag)++; ESP_LOGI("hrt", "START %d %s %s", id(tag), "GET", url.c_str());'
+        - http_request.get:
+            url: !lambda return url;
+            capture_response: true
+            max_response_buffer_size: ${bufsize}
+            on_response:
+              then:
+                - lambda: |-
+                    uint32_t sum = 0;
+                    for (char c : body) sum += (uint8_t) c;
+                    ESP_LOGI("hrt", "RESULT %d status=%d len=%u sum=%u took=%u maxgap=%u head=%.20s", id(tag),
+                             response->status_code, (unsigned) body.size(), (unsigned) sum,
+                             (unsigned) (millis() - id(t_start)), (unsigned) std::max<uint32_t>(id(max_gap), millis() - id(last_tick)), body.c_str());
+            on_error:
+              then:
+                - lambda: 'ESP_LOGI("hrt", "RESULT %d error took=%u maxgap=%u", id(tag), (unsigned) (millis() - id(t_start)), (unsigned) std::max<uint32_t>(id(max_gap), millis() - id(last_tick)));'
+        - lambda: 'ESP_LOGI("hrt", "CHAIN %d after send", id(tag));'
+    - action: req_post
+      variables:
+        url: string
+        payload: string
+      then:
+        - lambda: 'id(max_gap) = 0; id(t_start) = millis(); id(tag)++; ESP_LOGI("hrt", "START %d %s %s", id(tag), "POST", url.c_str());'
+        - http_request.post:
+            url: !lambda return url;
+            body: !lambda return payload;
+            capture_response: true
+            max_response_buffer_size: ${bufsize}
+            on_response:
+              then:
+                - lambda: |-
+                    uint32_t sum = 0;
+                    for (char c : body) sum += (uint8_t) c;
+                    ESP_LOGI("hrt", "RESULT %d status=%d len=%u sum=%u took=%u maxgap=%u head=%.20s", id(tag),
+                             response->status_code, (unsigned) body.size(), (unsigned) sum,
+                             (unsigned) (millis() - id(t_start)), (unsigned) std::max<uint32_t>(id(max_gap), millis() - id(last_tick)), body.c_str());
+            on_error:
+              then:
+                - lambda: 'ESP_LOGI("hrt", "RESULT %d error took=%u maxgap=%u", id(tag), (unsigned) (millis() - id(t_start)), (unsigned) std::max<uint32_t>(id(max_gap), millis() - id(last_tick)));'
+        - lambda: 'ESP_LOGI("hrt", "CHAIN %d after send", id(tag));'
+    - action: burst
+      variables:
+        url: string
+        count: int
+      then:
+        - lambda: 'id(max_gap) = 0; id(t_start) = millis(); ESP_LOGI("hrt", "BURST %d", count);'
+        - repeat:
+            count: !lambda return count;
+            then:
+              - script.execute:
+                  id: burst_script
+                  url: !lambda return url;
+    - action: cancel
+      variables:
+        url: string
+      then:
+        - lambda: 'id(max_gap) = 0; id(t_start) = millis(); ESP_LOGI("hrt", "CANCEL start");'
+        - script.execute:
+            id: single_script
+            url: !lambda return url;
+        - delay: 300ms
+        - script.execute:
+            id: single_script
+            url: !lambda return url;
+        - delay: 300ms
+        - script.stop: single_script
+        - lambda: 'ESP_LOGI("hrt", "CANCEL stopped running=%d", id(single_script).is_running());'
+        - delay: 4s
+        - lambda: 'ESP_LOGI("hrt", "CANCEL done maxgap=%u", (unsigned) std::max<uint32_t>(id(max_gap), millis() - id(last_tick)));'
+
+    - action: reboot_during
+      variables:
+        url: string
+      then:
+        - lambda: 'ESP_LOGI("hrt", "REBOOT_DURING start");'
+        - script.execute:
+            id: single_script
+            url: !lambda return url;
+        - script.execute:
+            id: burst_script
+            url: !lambda return url;
+        - delay: 500ms
+        - lambda: 'ESP_LOGI("hrt", "REBOOT_DURING rebooting"); App.safe_reboot();'
+
+script:
+  - id: burst_script
+    mode: parallel
+    max_runs: 10
+    parameters:
+      url: string
+    then:
+      - http_request.get:
+          url: !lambda return url;
+          capture_response: true
+          max_response_buffer_size: 256
+          on_response:
+            then:
+              - lambda: 'ESP_LOGI("hrt", "BURSTRESP status=%d len=%u took=%u maxgap=%u", response->status_code, (unsigned) body.size(), (unsigned) (millis() - id(t_start)), (unsigned) std::max<uint32_t>(id(max_gap), millis() - id(last_tick)));'
+          on_error:
+            then:
+              - lambda: 'ESP_LOGI("hrt", "BURSTERR took=%u", (unsigned) (millis() - id(t_start)));'
+  - id: single_script
+    parameters:
+      url: string
+    then:
+      - lambda: 'ESP_LOGI("hrt", "SINGLE begin");'
+      - http_request.get:
+          url: !lambda return url;
+          capture_response: true
+          max_response_buffer_size: 256
+          on_response:
+            then:
+              - lambda: 'ESP_LOGI("hrt", "SINGLERESP status=%d took=%u", response->status_code, (unsigned) (millis() - id(t_start)));'
+          on_error:
+            then:
+              - lambda: 'ESP_LOGI("hrt", "SINGLEERR took=%u", (unsigned) (millis() - id(t_start)));'
+      - lambda: 'ESP_LOGI("hrt", "SINGLE end");'
+
+http_request:
+  timeout: 8s
+
+globals:
+  - id: max_gap
+    type: uint32_t
+    initial_value: "0"
+  - id: last_tick
+    type: uint32_t
+    initial_value: "0"
+  - id: t_start
+    type: uint32_t
+    initial_value: "0"
+  - id: tag
+    type: int
+    initial_value: "0"
+
+interval:
+  - interval: 1ms
+    then:
+      - lambda: |-
+          const uint32_t now = millis();
+          if (id(last_tick) != 0 && now - id(last_tick) > id(max_gap))
+            id(max_gap) = now - id(last_tick);
+          id(last_tick) = now;
+
+ota:
+  - platform: esphome

.agents/skills/hardware-test/templates/http_request_action/hr32.yaml

@@ -0,0 +1,7 @@
+# Run file for the PR build; for the base build pass `-s name hr32base` to esphome (or copy it with that name).
+substitutions: {name: hr32pr, bufsize: "32768"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+logger:
+packages: {hr: !include hr-common.yaml}

.agents/skills/hardware-test/templates/http_request_action/hr66.yaml

@@ -0,0 +1,8 @@
+# Run file for the PR build; for the base build pass `-s name hr66base` to esphome (or copy it with that name).
+substitutions: {name: hr66pr, bufsize: "8192"}
+esp8266:
+  board: d1_mini
+logger:
+packages: {hr: !include hr-common.yaml}
+http_request:
+  verify_ssl: false

.agents/skills/hardware-test/templates/http_request_action/hrota32.yaml

@@ -0,0 +1,22 @@
+# Same as hr32.yaml plus the otaflash action otadrive.py calls; pass -s name hrota32base for the base build.
+substitutions: {name: hrota32, bufsize: "32768"}
+esp32:
+  board: esp32dev
+  framework: {type: esp-idf}
+logger:
+packages: {hr: !include hr-common.yaml}
+ota:
+  - platform: http_request
+    id: ota_http
+api:
+  actions:
+    - action: otaflash
+      variables:
+        url: string
+        sum: string
+      then:
+        - lambda: 'ESP_LOGI("hrt", "OTAFLASH start %s", url.c_str());'
+        - ota.http_request.flash:
+            url: !lambda return url;
+            md5: !lambda return sum;
+        - lambda: 'ESP_LOGI("hrt", "OTAFLASH returned (failed)");'

.agents/skills/hardware-test/templates/http_request_update/boards/upd32.yaml

@@ -0,0 +1,10 @@
+esphome:
+  name: upd32
+
+esp32:
+  board: esp32dev
+  framework:
+    type: esp-idf
+
+packages:
+  common: !include ../upd-common.yaml

.agents/skills/hardware-test/templates/http_request_update/boards/upd8266.yaml

@@ -0,0 +1,8 @@
+esphome:
+  name: upd8266
+
+esp8266:
+  board: d1_mini
+
+packages:
+  common: !include ../upd-common.yaml

.agents/skills/hardware-test/templates/http_request_update/upd-base-common.yaml

@@ -0,0 +1,52 @@
+substitutions:
+  base: !secret manifest_base
+
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+logger:
+  level: DEBUG
+api:
+  reboot_timeout: 0s
+
+http_request:
+  verify_ssl: false
+  timeout: 10s
+
+ota:
+  - platform: http_request
+    id: ota_http
+
+update:
+  - platform: http_request
+    id: upd
+    name: Firmware
+    source: ${base}/stable.json
+    update_interval: 1h
+
+button:
+  - platform: template
+    name: Check
+    on_press:
+      - lambda: id(upd).check();
+  - platform: template
+    name: URL Slow Cstr
+    on_press:
+      - lambda: |-
+          std::string url = "${base}/slow.json";
+          id(upd).set_source_url(url.c_str());
+  - platform: template
+    name: URL Stable Literal
+    on_press:
+      - lambda: id(upd).set_source_url("${base}/stable.json");
+  - platform: template
+    name: Heap
+    on_press:
+      - lambda: |-
+          #ifdef USE_ESP8266
+          ESP_LOGI("harness", "heap free %u", ESP.getFreeHeap());
+          #else
+          ESP_LOGI("harness", "heap free %u", (unsigned) heap_caps_get_free_size(MALLOC_CAP_INTERNAL));
+          #endif

.agents/skills/hardware-test/templates/http_request_update/upd-common.yaml

@@ -0,0 +1,95 @@
+substitutions:
+  base: !secret manifest_base
+
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+logger:
+  level: DEBUG
+  on_message:
+    - level: DEBUG
+      then:
+        - lambda: |-
+            // Inline check path: replace the URL from inside the check (ESP8266 runs the check on the main loop)
+            if (id(swap_in_check) && strstr(message, "Content-Length") != nullptr) {
+              id(swap_in_check) = false;
+              std::string url = "${base}/beta.json";
+              id(upd).set_source_url(url);
+              ESP_LOGI("harness", "URL replaced from inside the check");
+            }
+
+api:
+  reboot_timeout: 0s
+
+http_request:
+  verify_ssl: false
+  timeout: 10s
+
+ota:
+  - platform: http_request
+    id: ota_http
+
+update:
+  - platform: http_request
+    id: upd
+    name: Firmware
+    source: ${base}/stable.json
+    update_interval: 1h
+
+globals:
+  - id: swap_in_check
+    type: bool
+    initial_value: "false"
+
+button:
+  - platform: template
+    name: Check
+    on_press:
+      - lambda: id(upd).check();
+  - platform: template
+    name: URL Beta String
+    on_press:
+      - lambda: |-
+          std::string url = "${base}/beta.json";
+          id(upd).set_source_url(url);
+  - platform: template
+    name: URL Slow Cstr
+    on_press:
+      - lambda: |-
+          std::string url = "${base}/slow.json";
+          id(upd).set_source_url(url.c_str());
+  - platform: template
+    name: URL Stable Literal
+    on_press:
+      - lambda: id(upd).set_source_url("${base}/stable.json");
+  - platform: template
+    name: URL Temporary
+    on_press:
+      - lambda: id(upd).set_source_url(std::string("${base}/beta.json"));
+  - platform: template
+    name: Same URL Back
+    on_press:
+      - lambda: id(upd).set_source_url(id(upd).get_source_url());
+  - platform: template
+    name: Check Then Swap
+    on_press:
+      - lambda: |-
+          // ESP32: the check task reads slow.json while the URL is replaced under it
+          id(upd).check();
+          std::string url = "${base}/beta.json";
+          id(upd).set_source_url(url);
+  - platform: template
+    name: Arm Swap In Check
+    on_press:
+      - lambda: id(swap_in_check) = true;
+  - platform: template
+    name: Heap
+    on_press:
+      - lambda: |-
+          #ifdef USE_ESP8266
+          ESP_LOGI("harness", "heap free %u", ESP.getFreeHeap());
+          #else
+          ESP_LOGI("harness", "heap free %u", (unsigned) heap_caps_get_free_size(MALLOC_CAP_INTERNAL));
+          #endif

.agents/skills/hardware-test/templates/modem/modem-common.yaml

@@ -0,0 +1,65 @@
+esp32:
+  board: esp32dev
+  framework:
+    type: esp-idf
+  watchdog_timeout: 30s
+
+logger:
+  level: DEBUG
+
+api:
+  reboot_timeout: 0s
+
+modem:
+  id: atmodem
+  rx_pin: GPIO26
+  tx_pin: GPIO27
+  baud_rate: 115200
+  model: SIM7600
+  apn: internet
+  pin_code: "0000"
+  init_at:
+    - ATI
+  urc:
+    name: Modem URC
+  reboot_timeout: ${modem_reboot_timeout}
+  on_connect:
+    logger.log: "modem connected"
+  on_disconnect:
+    logger.log: "modem disconnected"
+  on_sync:
+    logger.log: "modem synced"
+
+sensor:
+  - platform: uptime
+    name: Uptime
+    update_interval: 30s
+  - platform: debug
+    free:
+      name: Heap Free
+    loop_time:
+      name: Loop Time
+
+debug:
+  update_interval: 30s
+
+button:
+  - platform: template
+    name: Modem Enable
+    on_press:
+      - modem.enable
+  - platform: template
+    name: Modem Disable
+    on_press:
+      - modem.disable
+  - platform: template
+    name: Modem Send AT
+    on_press:
+      - modem.send_at: "AT+CGSN"
+
+interval:
+  - interval: 20s
+    then:
+      - lambda: |-
+          ESP_LOGI("alive", "uptime %u s enabled=%d connected=%d", (unsigned) (millis() / 1000),
+                   id(atmodem).is_enabled(), id(atmodem).is_connected());

.agents/skills/hardware-test/templates/modem/modemo.yaml

@@ -0,0 +1,26 @@
+substitutions:
+  modem_reboot_timeout: 2min
+
+esphome:
+  name: modemo
+  on_boot:
+    priority: -100
+    then:
+      - modem.disable
+      - modem.enable
+      - if:
+          condition:
+            modem.enabled:
+          then:
+            modem.send_at: "AT+CGSN"
+          else:
+            modem.enable:
+      - if:
+          condition:
+            not:
+              modem.connected:
+          then:
+            logger.log: modem not connected at boot
+
+packages:
+  common: !include modem-common.yaml

.agents/skills/hardware-test/templates/modem/modemw.yaml

@@ -0,0 +1,12 @@
+substitutions:
+  modem_reboot_timeout: 0s
+
+esphome:
+  name: modemw
+
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+
+packages:
+  common: !include modem-common.yaml

.agents/skills/hardware-test/templates/online-image-s3-eth-dut.yaml

@@ -0,0 +1,105 @@
+esphome:
+  name: oi-s3eth
+  friendly_name: oi-s3eth
+
+esp32:
+  board: esp32-s3-devkitc-1
+  variant: esp32s3
+  flash_size: 16MB
+  framework:
+    type: esp-idf
+
+psram:
+  mode: octal
+  speed: 80MHz
+
+logger:
+  level: DEBUG
+  hardware_uart: USB_SERIAL_JTAG
+
+ethernet:
+  type: W5500
+  clk_pin: GPIO13
+  mosi_pin: GPIO11
+  miso_pin: GPIO12
+  cs_pin: GPIO14
+  interrupt_pin: GPIO10
+  reset_pin: GPIO9
+
+api:
+  encryption:
+    key: !secret api_key
+  reboot_timeout: 0s
+  actions:
+    - action: fetch
+      variables:
+        url: string
+      then:
+        - online_image.set_url:
+            id: img
+            url: !lambda 'return url;'
+            update: false
+        - component.update: img
+    - action: refetch
+      then:
+        - component.update: img
+    - action: fetch_small
+      variables:
+        url: string
+      then:
+        - online_image.set_url:
+            id: img_small
+            url: !lambda 'return url;'
+            update: false
+        - component.update: img_small
+    - action: refetch_small
+      then:
+        - component.update: img_small
+
+http_request:
+  timeout: 4.5s
+  verify_ssl: false
+
+display:
+
+image:
+  - platform: online_image
+    id: img
+    url: http://127.0.0.1/none.bmp
+    format: AUTO
+    type: RGB565
+    update_interval: never
+    on_download_finished:
+      logger.log:
+        format: "download finished: img cached=%d"
+        args: [cached]
+    on_error:
+      logger.log: "download error: img"
+  - platform: online_image
+    id: img_small
+    url: http://127.0.0.1/none.bmp
+    format: AUTO
+    type: RGB565
+    buffer_size: 256
+    update_interval: never
+    on_download_finished:
+      logger.log:
+        format: "download finished: img_small cached=%d"
+        args: [cached]
+    on_error:
+      logger.log: "download error: img_small"
+
+button:
+  - platform: restart
+    id: reboot_btn
+    name: Reboot
+
+debug:
+  update_interval: 2s
+
+sensor:
+  - platform: debug
+    loop_time:
+      name: Loop Time
+    free:
+      name: Heap Free

.agents/skills/hardware-test/templates/remote_base/rb-common.yaml

@@ -0,0 +1,33 @@
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+logger:
+  level: DEBUG
+
+api:
+  reboot_timeout: 0s
+  actions:
+    - action: tx_all
+      then:
+        - script.execute: tx_all
+
+ota:
+  - platform: esphome
+
+esphome:
+  on_boot:
+    priority: -100
+    then:
+      - lambda: |-
+          // A dumper registered at runtime, the way external components do it from setup()
+          struct ExtDumper : remote_base::RemoteReceiverDumperBase {
+            bool dump(remote_base::RemoteReceiveData src) override {
+              ESP_LOGI("rbt", "ext dumper saw %d", (int) src.size());
+              return false;
+            }
+          };
+          static ExtDumper ext;
+          id(rx_all).register_dumper(&ext);
+          ESP_LOGI("rbt", "ext dumper register called");

.agents/skills/hardware-test/templates/remote_base/rb-txscript.yaml

@@ -0,0 +1,46 @@
+script:
+  - id: tx_all
+    then:
+      - logger.log: "tx nec"
+      - remote_transmitter.transmit_nec: {address: 0x1234, command: 0x78}
+      - delay: 400ms
+      - logger.log: "tx samsung"
+      - remote_transmitter.transmit_samsung: {data: 0xE0E040BF}
+      - delay: 400ms
+      - logger.log: "tx sony"
+      - remote_transmitter.transmit_sony: {data: 0xA90, nbits: 12}
+      - delay: 400ms
+      - logger.log: "tx lg"
+      - remote_transmitter.transmit_lg: {data: 0x20DF10EF, nbits: 32}
+      - delay: 400ms
+      - logger.log: "tx jvc"
+      - remote_transmitter.transmit_jvc: {data: 0xC5E8}
+      - delay: 400ms
+      - logger.log: "tx panasonic"
+      - remote_transmitter.transmit_panasonic: {address: 0x4004, command: 0x0100BCBD}
+      - delay: 400ms
+      - logger.log: "tx rc5"
+      - remote_transmitter.transmit_rc5: {address: 0x1C, command: 0x05}
+      - delay: 400ms
+      - logger.log: "tx rc6"
+      - remote_transmitter.transmit_rc6: {address: 0x00, command: 0x0C}
+      - delay: 400ms
+      - logger.log: "tx dish"
+      - remote_transmitter.transmit_dish: {address: 1, command: 16}
+      - delay: 400ms
+      - logger.log: "tx coolix"
+      - remote_transmitter.transmit_coolix: {first: 0xB23FE4}
+      - delay: 400ms
+      - logger.log: "tx pioneer"
+      - remote_transmitter.transmit_pioneer: {rc_code_1: 0xA556}
+      - delay: 400ms
+      - logger.log: "tx samsung36"
+      - remote_transmitter.transmit_samsung36: {address: 0x0400, command: 0x000E00FF}
+      - delay: 400ms
+      - logger.log: "tx magiquest"
+      - remote_transmitter.transmit_magiquest: {wand_id: 0x01234567}
+      - delay: 400ms
+      - logger.log: "tx raw"
+      - remote_transmitter.transmit_raw: {code: [3000, -1000, 700, -2900, 1500, -400, 2200, -1700]}
+      - delay: 400ms
+      - logger.log: "tx done"

.agents/skills/hardware-test/templates/remote_base/rb32.tmpl

@@ -0,0 +1,32 @@
+esphome:
+  name: rb32@V@
+
+esp32:
+  board: esp32dev
+  framework:
+    type: esp-idf
+
+packages:
+  common: !include ../rb-common.yaml
+  txscript: !include ../rb-txscript.yaml
+
+remote_transmitter:
+  pin:
+    number: GPIO4
+    allow_other_uses: true
+  carrier_duty_percent: 100%
+  rmt_symbols: 64
+
+remote_receiver:
+  - id: rx_all
+    pin:
+      number: GPIO4
+      allow_other_uses: true
+    dump: all
+    rmt_symbols: 192
+  - id: rx_nec_raw
+    pin:
+      number: GPIO4
+      allow_other_uses: true
+    dump: [nec, raw]
+    rmt_symbols: 192

.agents/skills/hardware-test/templates/remote_base/rb66.tmpl

@@ -0,0 +1,69 @@
+esphome:
+  name: rb66@V@
+  includes:
+    - ../softtx.hpp
+
+esp8266:
+  board: d1_mini
+
+packages:
+  common: !include ../rb-common.yaml
+
+remote_receiver:
+  - id: rx_all
+    pin: GPIO5
+    dump: all
+
+# the receiver leaves its pin an input; drive it so the software sender reaches the pad
+interval:
+  - interval: 1h
+    startup_delay: 0s
+    then:
+      - lambda: pinMode(5, OUTPUT);
+
+script:
+  - id: tx_all
+    then:
+      - lambda: |-
+          remote_base::RemoteTransmitData d;
+          using namespace remote_base;
+          NECProtocol().encode(&d, {0x1234, 0x78, 1});
+          softtx::send(d, "nec");
+      - delay: 400ms
+      - lambda: |-
+          remote_base::RemoteTransmitData d;
+          using namespace remote_base;
+          SamsungProtocol().encode(&d, {0xE0E040BF, 32});
+          softtx::send(d, "samsung");
+      - delay: 400ms
+      - lambda: |-
+          remote_base::RemoteTransmitData d;
+          using namespace remote_base;
+          SonyProtocol().encode(&d, {0xA90, 12});
+          softtx::send(d, "sony");
+      - delay: 400ms
+      - lambda: |-
+          remote_base::RemoteTransmitData d;
+          using namespace remote_base;
+          LGProtocol().encode(&d, {0x20DF10EF, 32});
+          softtx::send(d, "lg");
+      - delay: 400ms
+      - lambda: |-
+          remote_base::RemoteTransmitData d;
+          using namespace remote_base;
+          JVCProtocol().encode(&d, {0xC5E8});
+          softtx::send(d, "jvc");
+      - delay: 400ms
+      - lambda: |-
+          remote_base::RemoteTransmitData d;
+          using namespace remote_base;
+          RC5Protocol().encode(&d, {0x1C, 0x05});
+          softtx::send(d, "rc5");
+      - delay: 400ms
+      - lambda: |-
+          remote_base::RemoteTransmitData d;
+          using namespace remote_base;
+          d.item(3000, 1000); d.item(700, 2900); d.item(1500, 400); d.mark(2200);
+          softtx::send(d, "raw");
+      - delay: 400ms
+      - logger.log: "tx done"

.agents/skills/hardware-test/templates/remote_base/softtx.hpp

@@ -0,0 +1,15 @@
+#pragma once
+// Bit-bang a frame with interrupts left on, so a receiver on the same pin sees every edge.
+// The ESP8266 remote_transmitter disables interrupts while it sends, which hides the edges.
+#include "esphome/components/remote_base/remote_base.h"
+
+namespace softtx {
+inline void send(const esphome::remote_base::RemoteTransmitData &d, const char *name) {
+  ESP_LOGI("rbt", "tx %s", name);
+  for (int32_t v : d.get_data()) {
+    digitalWrite(5, v > 0 ? HIGH : LOW);
+    delayMicroseconds(v > 0 ? v : -v);
+  }
+  digitalWrite(5, LOW);
+}
+}  // namespace softtx

.agents/skills/hardware-test/templates/store_yaml/boards/sy8266.yaml

@@ -0,0 +1,15 @@
+esphome:
+  name: sy8266
+
+esp8266:
+  board: d1_mini
+
+api:
+  encryption:
+    key: !secret api_key
+  reboot_timeout: 0s
+
+packages:
+  common: !include ../sy-common.yaml
+
+store_yaml:

.agents/skills/hardware-test/templates/store_yaml/boards/sy8266s.yaml

@@ -0,0 +1,16 @@
+esphome:
+  name: sy8266s
+
+esp8266:
+  board: d1_mini
+
+api:
+  encryption:
+    key: !secret api_key
+  reboot_timeout: 0s
+
+packages:
+  common: !include ../sy-common.yaml
+
+store_yaml:
+  include_secrets: true

.agents/skills/hardware-test/templates/store_yaml/sy-common.yaml

@@ -0,0 +1,24 @@
+# shared package for the store_yaml hardware test
+substitutions:
+  greeting: "hello from the package"
+
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+logger:
+  level: DEBUG
+
+ota:
+  - platform: esphome
+    password: recoverme123
+
+sensor:
+  - platform: template
+    name: Package Sensor
+    lambda: return 42.0;
+    update_interval: 60s
+
+packages:
+  extra: !include sy-extra.yaml

.agents/skills/hardware-test/templates/store_yaml/sy-extra.yaml

@@ -0,0 +1,6 @@
+# second level include
+text_sensor:
+  - platform: template
+    name: Greeting
+    lambda: return std::string("${greeting}");
+    update_interval: 60s

.agents/skills/hardware-test/templates/uart_sim/fr-common.yaml

@@ -0,0 +1,35 @@
+esp32:
+  board: esp32dev
+  framework:
+    type: esp-idf
+
+esphome:
+  includes:
+    - sim.hpp
+
+logger:
+  level: VERBOSE
+  logs:
+    component: DEBUG
+    api.service: DEBUG
+    api.connection: DEBUG
+    scheduler: DEBUG
+    sensor: DEBUG
+
+wifi:
+  ssid: !secret wifi_ssid
+  password: !secret wifi_password
+  power_save_mode: none
+
+api:
+  reboot_timeout: 0s
+
+debug:
+  update_interval: 10s
+
+sensor:
+  - platform: debug
+    free:
+      name: Heap Free
+    loop_time:
+      name: Loop Time

.agents/skills/hardware-test/templates/uart_sim/fra-parts.yaml

@@ -0,0 +1,86 @@
+esphome:
+  on_boot:
+    priority: -100
+    then:
+      - lambda: |-
+          sim::loopback(id(u_w));
+          sim::loopback(id(u_d));
+
+uart:
+  - id: u_w
+    tx_pin: GPIO25
+    rx_pin: GPIO26
+    baud_rate: 9600
+  - id: u_d
+    tx_pin: GPIO32
+    rx_pin: GPIO33
+    baud_rate: 9600
+
+sensor:
+  - platform: wts01
+    id: wts
+    name: WTS01 Temperature
+    uart_id: u_w
+    accuracy_decimals: 2
+    filters: []
+  - platform: ds1603l
+    name: DS1603L Distance
+    uart_id: u_d
+
+button:
+  - platform: template
+    name: W Good
+    on_press:
+      - lambda: sim::send(id(u_w), "W Good 22.18", sim::wts01(22, 18));
+  - platform: template
+    name: W Negative
+    on_press:
+      - lambda: sim::send(id(u_w), "W Negative -5.50", sim::wts01(0x85, 50));
+  - platform: template
+    name: W Extra 55
+    on_press:
+      - lambda: sim::send(id(u_w), "W Extra 55 then 23.25", sim::cat({0x55}, sim::wts01(23, 25)));
+  - platform: template
+    name: W 55 Breaks Header
+    on_press:
+      - lambda: sim::send(id(u_w), "W 55 01 01 then 26.00", sim::cat({0x55, 0x01, 0x01}, sim::wts01(26, 0)));
+  - platform: template
+    name: W Bad Checksum
+    on_press:
+      - lambda: sim::send(id(u_w), "W bad checksum 30.00", sim::wts01(30, 0, true));
+  - platform: template
+    name: W Garbage Then Frame
+    on_press:
+      - lambda: sim::send(id(u_w), "W garbage then 24.50", sim::cat({0x01, 0x02, 0x55, 0x01, 0x99}, sim::wts01(24, 50)));
+  - platform: template
+    name: W Split
+    on_press:
+      - lambda: sim::send(id(u_w), "W split head 27.75", sim::part(sim::wts01(27, 75), 0, 5));
+      - delay: 300ms
+      - lambda: sim::send(id(u_w), "W split rest", sim::part(sim::wts01(27, 75), 5, 9));
+  - platform: template
+    name: W Two Frames
+    on_press:
+      - lambda: sim::send(id(u_w), "W 20.01 then 20.02", sim::cat(sim::wts01(20, 1), sim::wts01(20, 2)));
+  - platform: template
+    name: D Good
+    on_press:
+      - lambda: sim::send(id(u_d), "D Good 1000", sim::ds1603l(1000));
+  - platform: template
+    name: D Misaligned
+    on_press:
+      - lambda: sim::send(id(u_d), "D FF 01 then 1234", sim::cat({0xFF, 0x01}, sim::ds1603l(1234)));
+  - platform: template
+    name: D Bad Checksum
+    on_press:
+      - lambda: sim::send(id(u_d), "D bad checksum 2000", sim::ds1603l(2000, true));
+  - platform: template
+    name: D Garbage Then Frame
+    on_press:
+      - lambda: sim::send(id(u_d), "D garbage then 500", sim::cat({0x12, 0x34}, sim::ds1603l(500)));
+  - platform: template
+    name: D Split
+    on_press:
+      - lambda: sim::send(id(u_d), "D split head 777", sim::part(sim::ds1603l(777), 0, 2));
+      - delay: 300ms
+      - lambda: sim::send(id(u_d), "D split rest", sim::part(sim::ds1603l(777), 2, 4));

.agents/skills/hardware-test/templates/uart_sim/fra.yaml

@@ -0,0 +1,6 @@
+# Run file for build a; copy it as frabase.yaml with name frabase for the base build.
+esphome:
+  name: fra
+packages:
+  common: !include fr-common.yaml
+  parts: !include fra-parts.yaml

.agents/skills/hardware-test/templates/uart_sim/frb-parts.yaml

@@ -0,0 +1,104 @@
+esphome:
+  on_boot:
+    priority: 200
+    then:
+      - lambda: |-
+          sim::loopback(id(u_s));
+          sim::loopback(id(u_f));
+
+uart:
+  - id: u_s
+    tx_pin: GPIO25
+    rx_pin: GPIO26
+    baud_rate: 9600
+  - id: u_f
+    tx_pin: GPIO32
+    rx_pin: GPIO33
+    baud_rate: 2400
+    parity: EVEN
+
+sensor:
+  - platform: sm300d2
+    id: sm
+    uart_id: u_s
+    update_interval: never
+    co2:
+      name: SM CO2
+    temperature:
+      name: SM Temperature
+  - platform: ufm01
+    temperature:
+      name: UFM Temperature
+    accumulated_flow:
+      name: UFM Accumulated Flow
+
+ufm01:
+  id: ufm
+  uart_id: u_f
+
+interval:
+  - interval: 1s
+    then:
+      - lambda: id(u_f).write_array(sim::ufm01(0x21, 0x50).data(), 32);
+
+script:
+  - id: sm_update
+    then:
+      - delay: 100ms
+      - lambda: |-
+          uint32_t t0 = micros();
+          id(sm).update();
+          ESP_LOGI("sim", "sm300d2 update() took %u us", (unsigned) (micros() - t0));
+
+button:
+  - platform: template
+    name: S Good
+    on_press:
+      - lambda: sim::send(id(u_s), "S Good co2 600 23.4", sim::sm300d2(600, 23, 4));
+      - script.execute: sm_update
+  - platform: template
+    name: S Empty Update
+    on_press:
+      - script.execute: sm_update
+  - platform: template
+    name: S Wrong Second Header
+    on_press:
+      - lambda: sim::send(id(u_s), "S 3C 05 then co2 700", sim::cat({0x3C, 0x05}, sim::sm300d2(700, 24, 0)));
+      - script.execute: sm_update
+  - platform: template
+    name: S Garbage Then Frame
+    on_press:
+      - lambda: sim::send(id(u_s), "S garbage then co2 800", sim::cat({0x01, 0x3C, 0x99}, sim::sm300d2(800, 25, 0)));
+      - script.execute: sm_update
+  - platform: template
+    name: S Cut Frame
+    on_press:
+      - lambda: sim::send(id(u_s), "S first 10 bytes of co2 900", sim::part(sim::sm300d2(900, 26, 0), 0, 10));
+      - script.execute: sm_update
+  - platform: template
+    name: S Two Frames
+    on_press:
+      - lambda: sim::send(id(u_s), "S co2 1001 then 1002", sim::cat(sim::sm300d2(1001, 27, 0), sim::sm300d2(1002, 28, 0)));
+      - script.execute: sm_update
+      - delay: 1s
+      - script.execute: sm_update
+  - platform: template
+    name: S Bad Checksum
+    on_press:
+      - lambda: |-
+          auto f = sim::sm300d2(1100, 29, 0);
+          f[16] += 0x11;
+          sim::send(id(u_s), "S bad checksum co2 1100", f);
+      - script.execute: sm_update
+  - platform: template
+    name: F Cut Then Frame
+    on_press:
+      - lambda: sim::send(id(u_f), "F 3C 32 00 00 then 33.33", sim::cat({0x3C, 0x32, 0x00, 0x00}, sim::ufm01(0x33, 0x33)));
+  - platform: template
+    name: F Bad Checksum Then Frame
+    on_press:
+      - lambda: sim::send(id(u_f), "F bad checksum 66.66 then 44.44", sim::cat(sim::ufm01(0x66, 0x66, true), sim::ufm01(0x44, 0x44)));
+  - platform: template
+    name: F Garbage Then Frame
+    on_press:
+      - lambda: sim::send(id(u_f), "F garbage then 55.55", sim::cat({0x16, 0x3C, 0x00}, sim::ufm01(0x55, 0x55)));