Users, groups, permissions, model settings and reasoning display

Four features, plus the schema machinery they needed.

**Schema sync.** The first live instance had data in it, and create_all
only creates missing *tables* -- a new column silently never appeared.
db/migrations.py now diffs the declared models against the database and
ALTER TABLE ... ADD COLUMN for what is missing, deriving a backfill
default from the column type (SQLite refuses a NOT NULL column without
one, and a Python-side `default=dict` cannot be expressed in DDL).
Verified against a copy of the live database: eight changes applied, all
rows preserved, second run a no-op. Renames, drops and retypes are still
manual and say so.

**Permissions.** A flat set of named booleans: an instance baseline
widened by each group the user belongs to. A group grants and never
denies -- with denies, "why can this user not do X" cannot be answered
without simulating every group. Admins bypass entirely, because an admin
can grant it back to themselves in two clicks and pretending otherwise
is theatre. Model *access* is separate: public, or granted to groups.
The picker is not the boundary -- switching a chat to a model you cannot
reach is a 403.

**Model settings.** Ordering, pinned-first, an instance default and a
per-user default, display names, descriptions, capability flags, and
uploaded images. Images are stored and served locally rather than by
URL: a remote URL makes every page render a request to a third party.
Uploads are validated by magic number, not the declared content type,
and stored under a random name. Models with no image get a generated
initial whose hue is derived from the model id, so it is stable.

**Reasoning display.** Streams into its own collapsible block above the
answer, labelled "Thought for 14 seconds", collapsed once finished, and
never replayed as context on the next turn. Two sources: the
reasoning_content delta field, and <think> tags inline in content -- the
latter needs a streaming splitter because the tags arrive split across
chunks. Models emitting no reasoning show nothing, via a :has() rule
rather than JavaScript. Verified against qwen35-9b on llama-swap: 694
reasoning events, 52 answer tokens, cleanly separated.

Two bugs found and fixed while testing:

- A bare `Mapped[list]` relationship is treated by SQLAlchemy as a scalar
  and returns None instead of []. It needs the element type.
- FastAPI substitutes the default for an empty form value, so with
  `x: str | None = Form(None)` a submitted `x=` is indistinguishable from
  an absent field. That silently broke clearing a system prompt or a
  temperature. update_chat now reads the raw form and checks key presence.

143 tests, ruff clean.

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

src/lembas/web/templates/chat/_message.html

@@ -39,6 +39,19 @@
     </header>
 
     {% if streaming %}
+      {# Reasoning arrives before the answer, so this block sits above it. It
+         starts open (watching a model think is the point) and the :has() rule
+         in chat.css hides the whole thing while it is still empty, so models
+         that emit no reasoning never show an empty box. #}
+      <details class="reasoning reasoning--live" id="reasoning-{{ message.id }}" open>
+        <summary class="reasoning__summary">
+          {{ icon("sparkle", "icon--sm reasoning__icon") }}
+          <span class="reasoning__label">Thinking</span>
+          {{ icon("chevron-down", "icon--sm reasoning__chevron") }}
+        </summary>
+        <div class="reasoning__body" sse-swap="reasoning" hx-swap="beforeend"></div>
+      </details>
+
       {# Tokens are appended here as they arrive. The cursor is a CSS
          pseudo-element on the empty parent, so it disappears by itself once
          the first token lands. #}
@@ -47,6 +60,25 @@
       <div class="msg__waiting">
         <span class="dots"><i></i><i></i><i></i></span>
       </div>
+    {% elif message.reasoning and not message.error %}
+      {# Collapsed once finished: the answer is what the reader came for, and
+         the thinking is there if they want to audit it. #}
+      <details class="reasoning" id="reasoning-{{ message.id }}">
+        <summary class="reasoning__summary">
+          {{ icon("sparkle", "icon--sm reasoning__icon") }}
+          <span class="reasoning__label">
+            {% if message.reasoning_ms %}
+              Thought for {{ message.reasoning_ms | duration }}
+            {% else %}
+              Reasoning
+            {% endif %}
+          </span>
+          {{ icon("chevron-down", "icon--sm reasoning__chevron") }}
+        </summary>
+        <div class="reasoning__body">{{ message.reasoning }}</div>
+      </details>
+      <div class="msg__body">{{ body_html|safe }}</div>
+
     {% elif message.error %}
       <div class="alert alert--error msg__error" role="alert">
         {{ icon("warning", "alert__icon") }}

src/lembas/web/templates/chat/index.html

@@ -23,22 +23,93 @@
       {% if chat %}
       <h1 class="topbar__title"><span id="chat-title">{{ chat.title }}</span></h1>
 
-      {% if models %}
-      <form hx-patch="/api/chats/{{ chat.id }}" hx-swap="none" hx-trigger="change from:find select">
+      {% if models and can.get("chat.model_select") %}
+      <form hx-patch="/api/chats/{{ chat.id }}" hx-swap="none"
+            hx-trigger="change from:find select">
         <select class="select select--compact" name="model_id" aria-label="Model">
-          {% for model in models %}
-          <option value="{{ model.model_id }}" {{ 'selected' if model.model_id == chat.model_id }}>
-            {{ model.label }}
-          </option>
+          {% if pinned_models %}
+          <optgroup label="Pinned">
+            {% for model in pinned_models %}
+            <option value="{{ model.model_id }}"
+                    {{ 'selected' if model.model_id == chat.model_id }}>{{ model.label }}</option>
+            {% endfor %}
+          </optgroup>
+          <optgroup label="Other models">
+          {% endif %}
+          {% for model in other_models %}
+          <option value="{{ model.model_id }}"
+                  {{ 'selected' if model.model_id == chat.model_id }}>{{ model.label }}</option>
           {% endfor %}
+          {% if pinned_models %}</optgroup>{% endif %}
         </select>
       </form>
+      {% elif current_model %}
+      <span class="badge">{{ current_model.label }}</span>
       {% endif %}
+
+      <button class="btn btn--icon" type="button" aria-label="Chat settings"
+              title="Chat settings"
+              onclick="document.getElementById('chat-settings').toggleAttribute('hidden')">
+        {{ icon("sliders") }}
+      </button>
       {% else %}
       <h1 class="topbar__title">Chats</h1>
       {% endif %}
     </header>
 
+    {% if chat and (can.get("chat.system_prompt") or can.get("chat.params")) %}
+    {# Collapsed by default: these are per-chat overrides, not everyday controls.
+       Each field saves on change rather than needing a Save button, so there is
+       no half-applied state to reason about. #}
+    <section class="chat-settings" id="chat-settings" hidden>
+      <div class="chat-settings__inner">
+        {% if current_model and current_model.description %}
+        <p class="chat-settings__note">{{ current_model.description }}</p>
+        {% endif %}
+
+        {% if can.get("chat.system_prompt") %}
+        <div class="field">
+          <label class="field__label" for="system-prompt">System prompt</label>
+          <textarea class="textarea" id="system-prompt" name="system_prompt" rows="3"
+                    placeholder="Instructions that apply to every message in this chat."
+                    hx-patch="/api/chats/{{ chat.id }}" hx-swap="none"
+                    hx-trigger="change">{{ chat.system_prompt }}</textarea>
+        </div>
+        {% endif %}
+
+        {% if can.get("chat.params") %}
+        <div class="chat-settings__params">
+          <div class="field">
+            <label class="field__label" for="temperature">Temperature</label>
+            <input class="input" id="temperature" name="temperature" type="number"
+                   min="0" max="2" step="0.05" placeholder="default"
+                   value="{{ chat.params_json.get('temperature', '') }}"
+                   hx-patch="/api/chats/{{ chat.id }}" hx-swap="none" hx-trigger="change">
+          </div>
+          <div class="field">
+            <label class="field__label" for="top-p">Top-p</label>
+            <input class="input" id="top-p" name="top_p" type="number"
+                   min="0" max="1" step="0.05" placeholder="default"
+                   value="{{ chat.params_json.get('top_p', '') }}"
+                   hx-patch="/api/chats/{{ chat.id }}" hx-swap="none" hx-trigger="change">
+          </div>
+          <div class="field">
+            <label class="field__label" for="max-tokens">Max tokens</label>
+            <input class="input" id="max-tokens" name="max_tokens" type="number"
+                   min="1" step="1" placeholder="default"
+                   value="{{ chat.params_json.get('max_tokens', '') }}"
+                   hx-patch="/api/chats/{{ chat.id }}" hx-swap="none" hx-trigger="change">
+          </div>
+        </div>
+        <p class="field__hint">
+          Leave a field empty to let the provider decide. Values outside the
+          allowed range are ignored rather than clamped.
+        </p>
+        {% endif %}
+      </div>
+    </section>
+    {% endif %}
+
     {% if not chat %}
       {# No chat selected. #}
       <div class="empty">