PWA, one send/stop button, audio in and out, web search as a tool

Four pieces of work.

**Installable.** A manifest carrying the instance name, PWA icons rasterised
from the existing mark at design time, a service worker and a themed offline
page. The worker caches the shell only and bails out on /api/, /auth/, /admin/
and anything accepting text/event-stream -- passing a reply stream through a
worker turns it into one delivery at the end, or nothing. It is served from
GET /sw.js rather than the static mount because a worker's scope is the path it
came from.

**Send and Stop are one button.** They were two, and the hidden one was never
hidden: `.btn` is display: inline-flex, which outranks the browser's own
`[hidden] { display: none }`, so Stop sat permanently beside Send. app.css now
forces the attribute to win -- every control toggled with `hidden` depended on
that -- and the composer renders one button carrying both icons, with ui.js
flipping data-composer-action and the type with it.

**Audio.** Speech to text and text to speech against any OpenAI-shaped
/v1/audio/* endpoint: dictate into the composer, have a reply read out.
Instance settings in Admin, per-reader overrides in Settings, with the voice
list discovered from the server where it offers one. Recorded audio is capped
and never written to disk -- it is not an attachment, it has no owner, and
nothing would ever sweep it.

**Web search, as a tool.** This is the tool loop PLAN.md described as the real
work: one reply is now a bounded sequence of requests rather than one. The model
asks, the tool runs, the result goes back and it is asked again, up to three
rounds. Providers are DuckDuckGo (no setup), SearXNG and Firecrawl.

Two decisions worth stating. Tools are only offered to models flagged `tools`,
because an endpoint without support rejects the whole request rather than
ignoring the array -- the same reason images only reach models flagged
`vision`. And tool results are not replayed as context on the next turn, for the
same reasons reasoning is not: the answer already contains what the model made
of them, and replaying stale results into every later request wastes the window
and reliably sends a small model into a search loop. The sources stay visible in
the transcript instead.

Search results are untrusted third-party text and are treated as such: escaped,
and only http/https URLs rendered as links.

338 tests, ruff clean.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

src/lembas/web/templates/partials/_voice_options.html

@@ -0,0 +1,25 @@
+{#
+  The contents of a voice <select>.
+
+  A fragment rather than markup inside each settings page because the admin
+  form, a reader's own settings and the endpoint test all need the same list --
+  which is fetched from the speech endpoint rather than stored, since
+  reconfiguring the server changes what is on offer.
+
+  `instance_voice` is only passed by the user-facing page, where the first
+  option means "whatever the administrator chose"; on the admin page there is
+  no such fallback and the empty option means "let the endpoint decide".
+#}
+{% if instance_voice is defined and instance_voice %}
+<option value="">Instance default — {{ instance_voice }}</option>
+{% else %}
+<option value="">Endpoint default</option>
+{% endif %}
+{% for voice in voices %}
+<option value="{{ voice }}" {{ 'selected' if voice == selected }}>{{ voice }}</option>
+{% endfor %}
+{% if selected and selected not in voices %}
+{# The stored choice is no longer offered -- kept so saving the form does not
+   silently reset it to the default. #}
+<option value="{{ selected }}" selected>{{ selected }} (not currently offered)</option>
+{% endif %}