diff --git a/ARCHITECTURE.md b/ARCHITECTURE.md
index 11e69e1..35b9cd0 100644
--- a/ARCHITECTURE.md
+++ b/ARCHITECTURE.md
@@ -841,12 +841,12 @@ be compacted independently. Unused branches keep their full history.
**What to implement when:**
-Phase 1 (v0.6.x): Add `parent_id` to messages table. Backfill
+Phase 1: Add `parent_id` to messages table. Backfill
existing messages with linear parent chains. All new messages created
with proper `parent_id`. Frontend still renders linearly — no branch
UI yet. This is the "do it now while it's cheap" step.
-Phase 2 (v0.7.x+): Edit-and-resubmit creates siblings instead of
+Phase 2: Edit-and-resubmit creates siblings instead of
replacing. Regenerate creates sibling model responses. Branch indicator
`← 1/2 →` in the UI. `channel_cursors` table for tracking active
branch.
@@ -887,7 +887,7 @@ supports tool calling.
Priority is based on: dependency depth (what blocks other things),
solo-user value (Jeff is the first user), and complexity.
-### Phase 1: Channel Foundation + Organization (v0.6.x)
+### Phase 1: Channel Foundation + Organization
**chats → channels migration + message tree + Folders + Projects + banners**
- Rename `chats` → `channels`, add `type`, `channel_members`,
`channel_models`, `participant_type`/`participant_id` on messages.
@@ -907,25 +907,25 @@ solo-user value (Jeff is the first user), and complexity.
- This is the schema foundation everything else builds on.
Do it now while there's one user and a handful of chats.
-### Phase 2: Notes + Built-in Tools (v0.7.x)
+### Phase 2: Notes + Built-in Tools ✅
**Notes CRUD + note_* tools + tool execution framework**
- Enables the LLM to be genuinely useful beyond ephemeral chat.
- Requires: tool execution in completion handler (the plumbing for
_all_ tools).
- This phase builds the tool calling infrastructure that everything
else uses.
-- **Conversation forking UI (v0.7.x):** Edit-and-resubmit creates
+- **Conversation forking UI (done, 0.7.2):** Edit-and-resubmit creates
siblings instead of replacing. Regenerate creates sibling model
responses. Branch indicator `← 1/2 →` at fork points. Context
assembly uses active path, not full channel history.
-### Phase 3: Web Search + URL Fetch (v0.7.x)
+### Phase 3: Web Search + URL Fetch
**web_search + url_fetch tools**
- Sidecar or direct HTTP from backend.
- Paired with Notes: "research X and save findings to my notes."
- Paired with Tasks (future): automated research.
-### Phase 4: @mention Routing + Multi-participant (v0.7.x)
+### Phase 4: @mention Routing + Multi-participant
**@mention parsing + multi-model channels**
- The channel schema exists from Phase 1. This phase adds the
routing logic: scan messages for @mentions, resolve against
@@ -934,28 +934,28 @@ solo-user value (Jeff is the first user), and complexity.
- Enables: editor mode (multiple model roles), second opinions,
cross-model conversations.
-### Phase 5: Embeddings + Knowledge Bases (v0.8.x)
+### Phase 5: Embeddings + Knowledge Bases
**Embedding pipeline + pgvector + KB CRUD + kb_search tool**
- Depends on: tool execution framework (Phase 2).
- Notes get embedded too (once pipeline exists).
- Admin configures embedding model.
- KBs attach to channels via `channel_models.kb_ids`.
-### Phase 6: Compaction (v0.8.x)
+### Phase 6: Compaction
**Auto-compaction service**
- Depends on: utility LLM calling (same as task runner).
- Channel-scoped: compacts any channel that exceeds context threshold.
- Can be built alongside or after KB (similar backend pattern:
background job that calls an LLM).
-### Phase 7: Tasks (v0.9.x)
+### Phase 7: Tasks
**Scheduler + task runner + task_create tool**
- Creates `type: 'service'` channels with no human members.
- Depends on: completion handler, tool execution, notes, web search.
- The capstone: everything below it combined into autonomous agents.
- Admin controls for resource limits.
-### Phase 8: Auth Strategy + Roles/Teams + Permissions (v0.10.x)
+### Phase 8: Auth Strategy + Roles/Teams + Permissions
**Enterprise auth modes + RBAC beyond admin/user**
- Auth middleware strategy pattern: `AUTH_MODE` env var selects
`builtin`, `mtls`, or `oidc`. All three resolve to the same
diff --git a/BRANDING.md b/BRANDING.md
new file mode 100644
index 0000000..0dd8af0
--- /dev/null
+++ b/BRANDING.md
@@ -0,0 +1,309 @@
+# Branding — Volume Mount Contract
+
+**Version:** 0.7.3
+**Status:** Spec
+
+---
+
+## Overview
+
+Chat Switchboard supports white-label branding through a volume mount at
+`/branding/` on the frontend container. Deployers provide a ConfigMap (or
+bind mount) with their identity assets. The app reads a JSON config on
+startup, serves static assets at runtime, and falls back gracefully when
+no branding is mounted.
+
+Switchboard provides the hooks. Your brand lives in its own repo.
+
+---
+
+## Mount Path
+
+```
+/branding/ ← volume mount root (frontend container)
+├── branding.json ← config seed (required if mount exists)
+├── favicon.png ← tab/bookmark icon
+├── logo.png ← splash page hero, optional sidebar
+└── custom.css ← style overrides (power-user escape hatch)
+```
+
+All files are optional individually, but `branding.json` is expected if the
+mount exists. Missing files degrade gracefully — no broken images, no JS errors.
+
+---
+
+## branding.json — Config Seed
+
+```json
+{
+ "org_name": "Gobha.ai",
+ "tagline": "Something clever here",
+ "accent_color": "#4a9eff",
+ "logo": "logo.png",
+ "favicon": "favicon.png"
+}
+```
+
+**Fields:**
+
+| Field | Type | Default | Description |
+|----------------|--------|----------------------|---------------------------------------------|
+| `org_name` | string | `"Chat Switchboard"` | Displayed in splash, header, auth card, `` |
+| `tagline` | string | `"Multi-Model AI Chat"` | Splash page subtitle, auth footer |
+| `headline` | string | `null` | Splash hero headline (default preserved if null) |
+| `accent_color` | string | `"#4a9eff"` | Primary UI accent (hex) |
+| `logo` | string | `null` | Filename relative to `/branding/` |
+| `favicon` | string | `null` | Filename relative to `/branding/` |
+| `pills` | array | (Switchboard defaults) | Feature pills on splash. `[]` = hide. See below. |
+
+**Pills format:**
+
+```json
+"pills": [
+ { "icon": "⚡", "text": "Fast inference", "style": "accent" },
+ { "icon": "🔒", "text": "Zero trust", "style": "purple" },
+ { "icon": "🏢", "text": "Enterprise ready" }
+]
+```
+
+`style` is optional: `"accent"` uses the accent color, `"purple"` uses purple,
+omit for the default neutral pill. Set `"pills": []` to hide the section entirely.
+Omit the field to keep the stock Switchboard pills.
+
+**Lifecycle:**
+
+1. Frontend entrypoint (`docker-entrypoint-fe.sh`) reads `/branding/branding.json` on container start
+2. Values are injected into `index.html` as a `<script>` block: `window.__BRANDING__`
+3. The `branding` key is also upserted into `global_settings` via the backend on startup
+ (admin panel override layer — if admins change values in the UI, those take precedence
+ until the next cold deploy with a branding mount)
+4. Frontend `initBranding()` applies values from `window.__BRANDING__` (fast, no API call)
+ then overlays any DB overrides from `App.serverSettings.branding` (from public settings)
+
+**Resolution order:** `branding.json` (fast init) → DB `global_settings.branding` (override layer)
+
+---
+
+## favicon.png — Static Asset
+
+Served at `/branding/favicon.png` by nginx. The frontend `<link rel="icon">` is
+set dynamically by `initBranding()`:
+
+```js
+// If branding favicon exists, use it; otherwise keep built-in default
+document.querySelector('link[rel="icon"]').href = '/branding/favicon.png';
+```
+
+**Recommendations:**
+- PNG format (modern browsers prefer it over ICO)
+- 32×32 minimum, 192×192 recommended (covers PWA + high-DPI)
+- Transparent background works best with dark themes
+
+---
+
+## logo.png — Static Asset
+
+Served at `/branding/logo.png`. Used in:
+- Splash page hero (replaces the 🔀 emoji)
+- Sidebar header (optional, depends on size)
+
+**Recommendations:**
+- PNG with transparency
+- Max 512×512 (larger files are wasteful; displayed at ~80px on splash)
+- Aspect ratio: square or landscape (tall logos will be clamped)
+
+---
+
+## custom.css — Style Extension
+
+Loaded *after* the main stylesheet:
+
+```html
+<link rel="stylesheet" href="/branding/custom.css" onerror="this.remove()">
+```
+
+The `onerror` handler silently removes the tag if the file doesn't exist (no 404
+console noise in unbrandeded deployments).
+
+**What you can do:**
+- Override `--accent-color` and any other CSS custom property
+- Change fonts (`@import` or `@font-face` with files in `/branding/`)
+- Add background textures or patterns
+- Hide elements you don't want (`display: none`)
+- Override specific component styles
+
+**What you shouldn't do:**
+- Rely on internal class names that may change between versions
+- Override layout properties (`flex`, `grid`) unless you're testing against the current release
+- Import external resources (breaks airgapped deployments)
+
+**Example:**
+
+```css
+:root {
+ --accent-color: #e74c3c;
+ --bg-primary: #1a1a2e;
+}
+
+.splash-logo img {
+ border-radius: 50%;
+}
+```
+
+---
+
+## Frontend Touchpoints
+
+These are the DOM elements and CSS properties that branding affects:
+
+| Element / Property | Default Value | Branding Source |
+|--------------------------|----------------------------------|-----------------------|
+| `<title>` | `Chat Switchboard` | `org_name` |
+| `.brand-text` | `Chat Switchboard` | `org_name` |
+| `.hero-wordmark` | `Chat Switchboard` | `org_name` |
+| `.hero-headline` | `One interface. Every AI model.` | `headline` |
+| `.hero-sub` | (default description) | `tagline` |
+| `.splash-logo` | SVG switchboard icon | `logo.png` → `<img>` |
+| `link[rel="icon"]` | `favicon-32.png` | `favicon.png` |
+| `--accent-color` | `#4a9eff` | `accent_color` |
+| `.auth-card-header h2` | `Welcome back` | `Welcome to {org_name}` |
+| `.auth-card-header p` | `Sign in to continue to your workspace` | `Sign in to continue` |
+| `.auth-footer p` | `Self-hosted AI chat...` | `tagline` |
+| `.hero-features` | Switchboard feature pills | `pills` array or `[]` to hide |
+
+---
+
+## Nginx Configuration
+
+The frontend entrypoint adds a branding location block. For path-based deployments,
+the block is under `BASE_PATH` so Traefik routes requests to the correct pod:
+
+```nginx
+# Root deployment (no BASE_PATH):
+location /branding/ {
+ alias /branding/;
+ expires 1h;
+ add_header Cache-Control "public";
+ try_files $uri =404;
+}
+
+# Path-based deployment (e.g. /dev, /test):
+location ${BASE_PATH}/branding/ {
+ alias /branding/;
+ expires 1h;
+ add_header Cache-Control "public";
+ try_files $uri =404;
+}
+```
+
+Frontend JS resolves paths via `window.__BASE__ + '/branding/'` so URLs
+automatically include the environment prefix.
+
+Short cache (1h) so branding updates via ConfigMap rollout are picked up
+without requiring users to hard-refresh.
+
+---
+
+## K8s Deployment
+
+The frontend deployment mounts the branding ConfigMap:
+
+```yaml
+spec:
+ containers:
+ - name: frontend
+ volumeMounts:
+ - name: branding
+ mountPath: /branding
+ readOnly: true
+ volumes:
+ - name: branding
+ configMap:
+ name: switchboard-branding
+ optional: true # ← app works without it
+```
+
+The `optional: true` is critical — Switchboard deploys cleanly with zero
+branding config. The `switchboard-gobha-ai` repo (or any deployer's
+equivalent) creates this ConfigMap.
+
+---
+
+## Deployer Repo Structure (Example: switchboard-gobha-ai)
+
+```
+switchboard-gobha-ai/
+├── branding/
+│ ├── branding.json
+│ ├── favicon.png
+│ ├── logo.png
+│ └── custom.css
+├── k8s/
+│ └── configmap.yaml
+├── README.md
+└── .gitea/
+ └── workflows/
+ └── deploy.yaml
+```
+
+**configmap.yaml:**
+
+```yaml
+apiVersion: v1
+kind: ConfigMap
+metadata:
+ name: switchboard-branding
+ namespace: ${NAMESPACE}
+data:
+ branding.json: |
+ {
+ "org_name": "Gobha.ai",
+ "tagline": "Your tagline",
+ "accent_color": "#4a9eff"
+ }
+binaryData:
+ favicon.png: <base64-encoded>
+ logo.png: <base64-encoded>
+ custom.css: <base64-encoded-or-use-data>
+```
+
+Note: For binary files in ConfigMaps, use `binaryData` with base64 encoding.
+Alternatively, use a script that creates the ConfigMap from files:
+
+```sh
+kubectl create configmap switchboard-branding \
+ --from-file=branding/branding.json \
+ --from-file=branding/favicon.png \
+ --from-file=branding/logo.png \
+ --from-file=branding/custom.css \
+ --dry-run=client -o yaml | kubectl apply -f -
+```
+
+---
+
+## Backend Integration
+
+The backend participates in branding in two ways:
+
+1. **Startup seed** (optional, future): If the backend also mounts `/branding/`,
+ it can read `branding.json` and upsert into `global_settings` alongside the
+ admin bootstrap. This enables admin-panel overrides without redeployment.
+
+2. **Public settings**: The `branding` key is added to `publicSettingKeys`,
+ making it available to non-admin users via `GET /api/v1/settings/public`.
+
+For 0.7.3, the frontend reads branding from the static mount at init time.
+Backend DB seeding is a future enhancement for admin-panel editing.
+
+---
+
+## Graceful Degradation
+
+| Condition | Behavior |
+|----------------------------------|-----------------------------------------------|
+| No `/branding/` mount | All defaults. App looks like stock Switchboard |
+| Mount exists, no `branding.json` | Static assets served, no text/color overrides |
+| `branding.json` missing fields | Each field falls back to its default |
+| `logo` field set, file missing | `<img>` gets 404, `onerror` shows emoji fallback |
+| `custom.css` missing | `<link>` tag self-removes via `onerror` |
+| `favicon.png` missing | Built-in favicon remains |
diff --git a/Dockerfile b/Dockerfile
index 0a07d3c..1d4919d 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,5 +1,5 @@
# ============================================
-# Chat Switchboard v0.5 - Unified Dockerfile
+# Chat Switchboard — Unified Dockerfile
# ============================================
# Stage 1: Build Go backend
# Stage 2: Download JS vendor libs (marked, DOMPurify)
diff --git a/EXTENSIONS.md b/EXTENSIONS.md
index e495603..0ba8e26 100644
--- a/EXTENSIONS.md
+++ b/EXTENSIONS.md
@@ -2,7 +2,7 @@
**Version:** 0.1 draft
**Status:** Design
-**Applies to:** v0.6.x+
+**Applies to:** v0.7.x+
**Companion to:** ARCHITECTURE.md (core services + terminology)
---
@@ -388,7 +388,7 @@ The cost tracker might show different metrics in editor mode vs chat mode.
### 6.4 Jeff's Planned Modes
-**Chat Mode** (core — v0.5.x)
+**Chat Mode** (core)
- What exists today, plus tool calling and richer message types.
- The "default surface" that everything else builds on.
@@ -572,7 +572,7 @@ publish to the server when something needs persistence or notification.
## 9. Implementation Roadmap
-### Phase A: Foundation (v0.6.0)
+### Phase A: Foundation
- [ ] `extensions.js` — loader, registry, scoped context
- [ ] `Extensions.register()` API with permission enforcement
- [ ] Manifest format parser and validator
@@ -580,7 +580,7 @@ publish to the server when something needs persistence or notification.
- [ ] Asset serving endpoint
- [ ] Extension settings UI in Settings modal
-### Phase B: Browser Tools (v0.6.1)
+### Phase B: Browser Tools
- [ ] `ctx.tools.handle()` API for browser tool registration
- [ ] Tool schema collection in completion handler
- [ ] `tool.call.*` / `tool.result.*` WebSocket routing
@@ -588,19 +588,19 @@ publish to the server when something needs persistence or notification.
- [ ] Tool execution in EventBus `WaitFor` pattern
- [ ] Tool use display in chat messages
-### Phase C: Surfaces (v0.6.2)
+### Phase C: Surfaces
- [ ] Surface registration and activation API
- [ ] Mode selector in sidebar
- [ ] `ctx.ui.replace()` / `ctx.ui.restore()` for region management
- [ ] `surface.activated` / `surface.deactivated` events
-### Phase D: First Extensions (v0.7.0)
+### Phase D: First Extensions
- [ ] Cost tracker (browser, proof of concept)
- [ ] Slash commands (browser, message transforms)
- [ ] Custom renderers (browser, mermaid/latex/etc)
- [ ] Editor mode (browser, with git provider tools)
-### Phase E: Server-Side Tiers (v0.8.0)
+### Phase E: Server-Side Tiers
- [ ] Starlark runtime integration
- [ ] Sidecar HTTP tool protocol
- [ ] Server-side tool execution in completion handler
diff --git a/FORKING_IMPL.md b/FORKING_IMPL.md
deleted file mode 100644
index 5de14f3..0000000
--- a/FORKING_IMPL.md
+++ /dev/null
@@ -1,96 +0,0 @@
-# Message Editing & Forking — Implementation
-
-**Version:** 0.7.x (Phase 2 per ARCHITECTURE.md §8)
-**Depends on:** parent_id (006), channel_cursors (008) — both deployed
-
----
-
-## Status: Complete (backend + frontend)
-
-### Backend (already existed)
-- [x] Migration 013 (`deleted_at`, `sibling_index`)
-- [x] `tree.go` — `getActivePath`, `getActiveLeaf`, `getPathToLeaf`, `getSiblings`,
- `findLeafFromMessage`, `nextSiblingIndex`, `updateCursor`
-- [x] `completion.go` — `loadConversation` walks active path via recursive CTE,
- `persistMessage` is cursor-aware with sibling_index, returns new message ID
-- [x] `messages.go` — `GetActivePath`, `EditMessage`, `Regenerate` (streaming + sibling),
- `UpdateCursor`, `ListSiblings`
-- [x] Routes wired in `main.go`
-
-### Frontend (new work)
-- [x] `api.js` — `getActivePath`, `editMessage`, `streamRegenerate`, `updateCursor`, `listSiblings`
-- [x] `app.js` — tree-aware message flow:
- - `selectChat` uses `/path` endpoint
- - `sendMessage` reloads path after streaming for server-authoritative IDs
- - `regenerateMessage(id)` creates sibling via message-specific endpoint
- - `editMessage(id)` → inline edit → `submitEdit` → sibling + completion
- - `switchSibling(id, direction)` navigates branches via cursor update
- - `reloadActivePath()` refreshes from server after every mutation
- - Legacy `regenerate()` button delegates to `regenerateMessage(lastAssistantId)`
-- [x] `ui.js`:
- - `_messageHTML` renders ‹ 1/3 › branch indicators at fork points
- - Edit button on user messages, Regen button on assistant messages
- - `showEditInline` — textarea with Ctrl+Enter submit, Escape cancel
-- [x] `styles.css` — `.branch-nav`, `.branch-arrow`, `.branch-pos`,
- `.msg-edit-input`, `.msg-edit-actions`, `.msg-edit-cancel`, `.msg-edit-submit`
-
----
-
-## API Endpoints
-
-| Method | Path | Purpose |
-|--------|------|---------|
-| `GET` | `/channels/:id/path` | Active path with sibling metadata |
-| `POST` | `/channels/:id/messages/:msgId/edit` | Create sibling user message |
-| `POST` | `/channels/:id/messages/:msgId/regenerate` | Create sibling assistant response (SSE) |
-| `PUT` | `/channels/:id/cursor` | Switch active branch |
-| `GET` | `/channels/:id/messages/:msgId/siblings` | List siblings at a fork point |
-
----
-
-## Design Decisions
-
-1. **`sibling_index` column** — explicit integer, not derived from `created_at`. Tree is self-describing.
-2. **New `/path` endpoint** — `ListMessages` retained for admin/debug. Frontend uses `/path` exclusively.
-3. **Two-step edit** — API: edit creates sibling, completion is separate. Frontend chains them.
-4. **Uniform ‹ › arrows** — same UI regardless of how fork was created (edit, regen, explicit).
-5. **Message-specific regenerate** — replaces old channel-level stub. Bottom-bar button delegates.
-6. **Branch switch → leaf** — clicking mid-tree sibling walks to deepest descendant via `findLeafFromMessage`.
-7. **Path reload after every mutation** — `reloadActivePath()` called post-send, edit, regen, abort.
-
----
-
-## User Flows
-
-### Edit a message
-1. Hover user message → click **Edit**
-2. Textarea replaces message text (Ctrl+Enter to submit, Escape to cancel)
-3. Submit → `POST /messages/:id/edit` (creates sibling) → `POST /chat/completions` (new response)
-4. Branch indicator appears: ‹ 1/2 ›
-5. Old branch preserved, navigable via arrows
-
-### Regenerate a response
-1. Hover assistant message → click **Regen** (or bottom-bar ↻ button)
-2. `POST /messages/:id/regenerate` streams new response as sibling
-3. Branch indicator appears: ‹ 1/2 ›
-4. Original response preserved, navigable via arrows
-
-### Navigate branches
-1. Click ‹ or › on any fork point
-2. `GET /messages/:id/siblings` → pick adjacent → `PUT /cursor` with sibling ID
-3. Backend walks to leaf, returns new path
-4. Entire conversation below fork point updates to show selected branch
-
----
-
-## Testing Checklist
-
-- [ ] Existing linear conversations load and render correctly (backward compat)
-- [ ] New messages in linear conversation work as before
-- [ ] Regen creates sibling, shows ‹ 1/2 ›, original preserved
-- [ ] Multiple regens show ‹ 1/3 ›, all navigable
-- [ ] Edit creates sibling user message + triggers new completion
-- [ ] Branch arrows navigate correctly (← older, → newer)
-- [ ] Context sent to LLM is active path only (no cross-branch pollution)
-- [ ] Cursor persists across page reload (reopening chat shows last-viewed branch)
-- [ ] Mobile: edit textarea usable, branch arrows tappable
diff --git a/ISSUES.md b/ISSUES.md
index 170ea20..138cea0 100644
--- a/ISSUES.md
+++ b/ISSUES.md
@@ -29,9 +29,8 @@ Longer explanation if needed. Closes #XX
## Current Priorities
-1. WebSocket hub (blocks Channels)
-2. Channels backend + frontend
-3. Plugin architecture design
-4. Notes & Knowledge Base
+1. UX polish (chat search, keyboard shortcuts, PWA)
+2. Teams + RBAC (0.8.0)
+3. Audit + Usage tracking (0.8.x)
See [ROADMAP.md](ROADMAP.md) for detail.
diff --git a/ROADMAP.md b/ROADMAP.md
index 2b4875e..187bcee 100644
--- a/ROADMAP.md
+++ b/ROADMAP.md
@@ -10,7 +10,7 @@ minor = add features (deprecate, don't remove), patch = fix something.
---
-## Current State: v0.7.0
+## Current State: v0.7.3
### ✅ Done
@@ -42,6 +42,12 @@ minor = add features (deprecate, don't remove), patch = fix something.
- [x] User + admin provider model listing with capabilities
- [x] Model presets (named wrappers: global/personal scope, system prompt, temp, max_tokens)
- [x] Preset unwrap in completion handler (transparent to provider)
+- [x] Tool framework: registry, type system, execution loop (tool→result→model)
+- [x] Note tools: note_create, note_search, note_update, note_list
+- [x] Notes API: CRUD, full-text search (tsvector), folder listing
+- [x] Provider tool calling: OpenAI + Anthropic function calling in request/response/stream
+- [x] Message tree: edit creates sibling, regenerate creates sibling, cursor tracking
+- [x] Avatar system: user + preset avatars, server-side resize to 128×128 PNG
**Frontend (Vanilla JS)**
- [x] Professional splash page (split-panel hero + tabbed auth)
@@ -68,6 +74,11 @@ minor = add features (deprecate, don't remove), patch = fix something.
- [x] Appearance settings tab (UI scale, message font size)
- [x] Mobile responsive layout (hamburger menu, sidebar overlay, dvh)
- [x] Model/preset name in message headers (replaces generic "Assistant")
+- [x] Notes modal: list/detail views, folder sidebar, create/edit/delete, search
+- [x] Message editing + forking: inline edit, regen, branch navigation ‹ 1/2 ›
+- [x] Conversation path: active-path context assembly, cursor-aware
+- [x] White-label branding: volume mount, org name, logo, favicon, accent color, pills
+- [x] Avatar system: user profile upload, preset avatars, avatar in messages/sidebar/dropdown
**CI/CD (Gitea Actions)**
- [x] Three-env pipeline (dev/test/prod)
@@ -106,26 +117,24 @@ org-wide presets, users create personal ones (if user providers are enabled).
White-label support and UX improvements that exploit existing infrastructure.
**Branding**
-- [ ] Admin branding settings in global_settings: org name, tagline, accent color
-- [ ] `/branding/` volume mount (K8s ConfigMap, optional) for favicon, logo
-- [ ] Splash page reads branding config on load, falls back to Switchboard defaults
-- [ ] CSS accent color override from branding settings
+- [x] Admin branding settings in global_settings: org name, tagline, accent color
+- [x] `/branding/` volume mount (K8s ConfigMap, optional) for favicon, logo
+- [x] Splash page reads branding config on load, falls back to Switchboard defaults
+- [x] CSS accent color override from branding settings
**Profile Pictures / Avatars**
-- [ ] `avatar` column on `users` table (base64 PNG, capped ~128x128, nullable)
-- [ ] `preset_avatar` column on `model_presets` (alongside existing `icon` emoji field)
-- [ ] Avatar upload in user Settings → Profile section (crop/resize client-side)
-- [ ] Admin preset form: optional avatar upload (or keep emoji-only)
-- [ ] `UI.avatar(msg)` helper: returns `<img>` if avatar set, emoji fallback otherwise
-- [ ] Replace hardcoded 👤/🤖 in `_messageHTML`, `streamResponse`, sidebar user area
-- [ ] Avatar displayed in message headers, chat list, user flyout
-- [ ] Storage: base64 in DB keeps backups self-contained (airgap-friendly)
+- [x] `avatar_url` column on `users` table (already existed, now wired)
+- [x] `avatar` column on `model_presets` (migration 015)
+- [x] Avatar upload in user Settings → Profile section (server-side resize to 128×128)
+- [x] Admin/user preset avatar upload endpoint
+- [x] `avatarHTML()` helper: returns `<img>` if avatar set, emoji fallback otherwise
+- [x] Avatar displayed in message headers, stream, typing indicator, sidebar user area
**Message Editing + Forking**
-- [ ] Edit message → creates sibling (uses existing parent_id tree)
-- [ ] Regenerate → creates sibling model response
-- [ ] Branch indicator at fork points (← 1/2 →)
-- [ ] Context assembly follows active path, not full channel history
+- [x] Edit message → creates sibling (uses existing parent_id tree)
+- [x] Regenerate → creates sibling model response
+- [x] Branch indicator at fork points (← 1/2 →)
+- [x] Context assembly follows active path, not full channel history
**UX Polish**
- [ ] Chat search / filter in sidebar
@@ -191,28 +200,28 @@ Required for enterprise and compliance. Cheap to build, expensive to retrofit.
---
-## 0.9.0 — Tool Execution + Notes
+## ~~0.9.0 — Tool Execution + Notes~~ ✅ (pulled into 0.7.x)
-The tool-calling infrastructure that everything downstream depends on.
+Pulled forward and shipped ahead of schedule. See TOOLS_IMPL.md and FORKING_IMPL.md.
**Tool Framework**
-- [ ] Tool calling pipeline in completion handler
-- [ ] Tool registry (built-in + future plugin tools)
-- [ ] Tool execution loop: model requests tool → backend executes → result fed back
+- [x] Tool calling pipeline in completion handler (OpenAI + Anthropic)
+- [x] Tool registry (built-in + future plugin tools)
+- [x] Tool execution loop: model requests tool → backend executes → result fed back
- [ ] Tool permission model (which tools enabled per preset/channel)
**Notes**
-- [ ] `notes` table: title, content, folder_id, tags, team_id (nullable), created_by
-- [ ] Notes CRUD endpoints
-- [ ] `note_create`, `note_update`, `note_search`, `note_list` tools
-- [ ] Full-text search (PostgreSQL `tsvector`)
-- [ ] Markdown editor in frontend
-- [ ] Team-scoped notes (schema ready from 0.8.0)
+- [x] `notes` table: title, content, folder, tags, source_channel_id, full-text search
+- [x] Notes CRUD endpoints + search
+- [x] `note_create`, `note_update`, `note_search`, `note_list` tools
+- [x] Full-text search (PostgreSQL `tsvector` + `ts_rank` + `ts_headline`)
+- [x] Notes UI: modal with list/detail views, folder sidebar, Markdown editor
+- [ ] Team-scoped notes (awaits 0.8.0 teams)
**Conversation Forking UI**
-- [ ] Edit-and-resubmit creates siblings (tree structure from 0.6.0 schema)
-- [ ] Branch indicator ← 1/2 → at fork points
-- [ ] Context assembly follows active path
+- [x] Edit-and-resubmit creates siblings (tree structure)
+- [x] Branch indicator ← 1/2 → at fork points
+- [x] Context assembly follows active path
## 0.9.x — Context Management
@@ -228,7 +237,7 @@ Usability before compaction exists — long conversations shouldn't silently fai
## 0.10.0 — Web Search + URL Fetch
-First real external tools, built on 0.9.0 framework.
+First external tools, using the tool framework shipped in 0.7.x.
- [ ] `web_search` tool: search provider abstraction (DuckDuckGo, SearXNG, Brave)
- [ ] `url_fetch` tool: retrieve and extract content from URLs
@@ -268,14 +277,14 @@ storage, metadata and search indexes live in PostgreSQL.
- [ ] Paste-to-upload (clipboard image support)
*Note: media generation (image gen, video models) is a separate concern —
-those are tool-use actions that depend on the tool framework (0.9.0) and
+those are tool-use actions that depend on the tool framework and
produce attachments as output. Tracked under Future.*
---
## 0.12.0 — @mention Routing + Multi-model
-The channel schema exists from 0.6.0. This phase adds the routing logic.
+The channel schema already supports multiple models. This phase adds the routing logic.
- [ ] @mention parsing in messages (users and AI models)
- [ ] Resolve mentions against `channel_models`
@@ -292,7 +301,7 @@ The channel schema exists from 0.6.0. This phase adds the routing logic.
- [ ] `knowledge_bases` table: name, description, team_id (nullable), created_by
- [ ] KB document storage via 0.11.0 storage backend (same S3/PVC abstraction)
- [ ] KB CRUD endpoints + admin UI
-- [ ] `kb_search` tool (built on 0.9.0 tool framework)
+- [ ] `kb_search` tool (uses existing tool framework)
- [ ] Context injection in completion flow
- [ ] Notes get embedded too (once pipeline exists)
- [ ] Team admins manage team KBs (permission layer from 0.8.0)
@@ -391,7 +400,7 @@ could pull left based on need.
## Extension / Plugin Architecture
-**Deferred to post-1.0.** The tool execution framework (0.9.0) provides the
+**Deferred to post-1.0.** The tool execution framework (0.7.x) provides the
internal hook points. A formal plugin API, manifest format, and marketplace
are tracked in the dedicated design documents:
diff --git a/TOOLS_IMPL.md b/TOOLS_IMPL.md
deleted file mode 100644
index 4bab1c1..0000000
--- a/TOOLS_IMPL.md
+++ /dev/null
@@ -1,102 +0,0 @@
-# Notes & Tool Execution Framework — Implementation
-
-**Version:** 0.7.2 (Phase 2 per ARCHITECTURE.md §10)
-**Depends on:** channels (006), messages with tree (013) — both deployed
-
----
-
-## Status: Complete
-
-### Migration
-- [x] `014_notes.sql` — notes table, tsvector full-text search, folder/tag indexes, auto-update trigger
-
-### Notes API (handlers/notes.go)
-- [x] `POST /notes` — create note with title, content, folder, tags, source_channel_id
-- [x] `GET /notes` — list with folder/tag filter, pagination
-- [x] `GET /notes/:id` — get single note
-- [x] `PUT /notes/:id` — update with replace/append/prepend modes
-- [x] `DELETE /notes/:id` — delete
-- [x] `GET /notes/search?q=` — full-text search with ts_rank + ts_headline
-- [x] `GET /notes/folders` — list distinct folder paths with counts
-
-### Tool Type System (tools/types.go)
-- [x] `ToolDef` — name, description, JSON Schema parameters (sent to LLM)
-- [x] `ToolCall` — ID, name, arguments (from LLM response)
-- [x] `ToolResult` — tool_call_id, name, content, is_error (fed back)
-- [x] `Tool` interface — `Definition()` + `Execute(ctx, execCtx, argsJSON)`
-- [x] `ExecutionContext` — userID, channelID for scoped execution
-- [x] JSON Schema helpers: `Prop`, `PropEnum`, `PropArray`, `JSONSchema`
-
-### Tool Registry (tools/registry.go)
-- [x] `Register(Tool)` — global registry, called from init()
-- [x] `Get(name)` — lookup by name
-- [x] `AllDefinitions()` — all registered tool schemas for LLM requests
-- [x] `ExecuteCall()` — runs a single tool, returns ToolResult (never errors)
-- [x] `ExecuteAll()` — batch execution
-- [x] `HasTools()` — check if any tools registered
-
-### Note Tools (tools/notes.go)
-- [x] `note_create` — create with title, content, folder, tags; links source_channel_id
-- [x] `note_search` — full-text search with relevance ranking and excerpts
-- [x] `note_update` — update by note_id with replace/append/prepend modes
-- [x] `note_list` — list by folder/tag filter
-- All registered via init()
-
-### Provider Tool Calling Support
-- [x] `providers.Message` — extended with ToolCalls, ToolCallID, Name fields
-- [x] `providers.CompletionRequest` — Tools field for function definitions
-- [x] `providers.CompletionResponse` — ToolCalls field for function calls
-- [x] `providers.StreamEvent` — ToolCalls accumulated on final event
-- [x] OpenAI provider — full tool_calls in request/response/stream
-- [x] Anthropic provider — tool_use/tool_result content blocks, input_json_delta streaming, consecutive user message merging
-- [x] OpenRouter — inherits from OpenAI (delegation)
-- [x] Venice — inherits from OpenAI (delegation)
-
-### Completion Handler Tool Loop (handlers/completion.go)
-- [x] `buildToolDefs()` — converts registered tools to provider format
-- [x] Tools attached when `caps.ToolCalling && tools.HasTools()`
-- [x] Streaming: tool loop with max 10 iterations
- - Stream text deltas normally
- - On `finish_reason: "tool_calls"`: execute tools, send SSE events
- - Custom SSE events: `event: tool_use` and `event: tool_result`
- - Re-call provider with tool results appended to messages
- - Final text response streamed normally
-- [x] Non-streaming: same loop, tools executed synchronously
-- [x] Safety: `maxToolIterations = 10` prevents infinite loops
-
----
-
-## SSE Events for Tool Execution
-
-The streaming completion sends custom named SSE events during tool execution.
-Standard data events continue to carry OpenAI-compatible content deltas.
-
-| Event | Payload | When |
-|-------|---------|------|
-| `tool_use` | `[{"id":"...","type":"function","function":{"name":"note_create","arguments":"..."}}]` | LLM requests tool calls |
-| `tool_result` | `{"tool_call_id":"...","name":"note_create","content":"...","is_error":false}` | Tool execution complete |
-
-The frontend can listen for these to show tool execution status.
-Existing frontends that only handle `data:` events are unaffected.
-
----
-
-## Design Decisions
-
-1. **Separate packages** — `tools/` is independent of `handlers/`. Tools don't know about HTTP. The completion handler bridges them.
-2. **init() registration** — Tools self-register. Adding a new tool = create file with init(), done. No central wiring.
-3. **JSON Schema parameters** — Tool definitions use standard JSON Schema, matching OpenAI's function calling spec.
-4. **Non-streaming tool loop** — After tool execution, re-calls use streaming. Each iteration streams text deltas.
-5. **Anthropic content blocks** — Full translation: `tool_use` blocks, `tool_result` as user messages, consecutive user message merging for alternating-role requirement.
-6. **Folder paths** — Normalized to `/path/` format. Root is `/`. Tree structure without a folders table.
-7. **Search** — PostgreSQL tsvector with weighted ranks (title=A, content=B). No external search infrastructure needed.
-
----
-
-## Frontend Work
-
-- [x] Notes panel/modal in UI
-- [x] Notes CRUD forms (create, edit, list, search)
-- [x] Tool execution indicators during streaming
-- [x] Parse `event: tool_use` / `event: tool_result` SSE events
-- [x] Display tool activity in message bubbles
diff --git a/VERSION b/VERSION
index 7486fdb..f38fc53 100644
--- a/VERSION
+++ b/VERSION
@@ -1 +1 @@
-0.7.2
+0.7.3
diff --git a/docker-entrypoint-fe.sh b/docker-entrypoint-fe.sh
index 0d2d18c..e785d96 100644
--- a/docker-entrypoint-fe.sh
+++ b/docker-entrypoint-fe.sh
@@ -26,11 +26,22 @@ else
BASE_HREF="${BASE_PATH}/"
fi
+# ── Read branding config ────────────────────
+BRANDING_JSON="{}"
+if [ -f /branding/branding.json ]; then
+ # Compact to single line for safe sed injection
+ BRANDING_JSON=$(tr -d '\n' < /branding/branding.json | sed 's/ */ /g')
+ echo "✅ Branding config loaded from /branding/branding.json"
+else
+ echo "ℹ️ No branding mount — using defaults"
+fi
+
# ── Inject into index.html ──────────────────
sed -i \
-e "s|%%BASE_PATH%%|${BASE_PATH}|g" \
-e "s|%%BASE_HREF%%|${BASE_HREF}|g" \
-e "s|%%APP_VERSION%%|${APP_VERSION}|g" \
+ -e "s|%%BRANDING_JSON%%|${BRANDING_JSON}|g" \
/usr/share/nginx/html/index.html
echo "✅ Frontend configured: BASE_PATH=${BASE_PATH:-/} VERSION=${APP_VERSION}"
@@ -57,6 +68,14 @@ server {
add_header Cache-Control "public, immutable";
}
+ # Branding assets — served from volume mount, 404s gracefully
+ location /branding/ {
+ alias /branding/;
+ expires 1h;
+ add_header Cache-Control "public";
+ try_files $uri =404;
+ }
+
location / {
try_files $uri $uri/ /index.html;
}
@@ -94,6 +113,14 @@ server {
}
}
+ # Branding assets — under BASE_PATH so Traefik routes them here
+ location ${BASE_PATH}/branding/ {
+ alias /branding/;
+ expires 1h;
+ add_header Cache-Control "public";
+ try_files \$uri =404;
+ }
+
# Redirect bare path to trailing slash
location = ${BASE_PATH} {
return 301 ${BASE_PATH}/;
diff --git a/k8s/frontend.yaml b/k8s/frontend.yaml
index 6fe3921..53a81ce 100644
--- a/k8s/frontend.yaml
+++ b/k8s/frontend.yaml
@@ -38,6 +38,10 @@ spec:
env:
- name: BASE_PATH
value: "${BASE_PATH}"
+ volumeMounts:
+ - name: branding
+ mountPath: /branding
+ readOnly: true
resources:
requests:
memory: "${FE_MEMORY_REQUEST}"
@@ -61,6 +65,11 @@ spec:
periodSeconds: 30
timeoutSeconds: 5
failureThreshold: 3
+ volumes:
+ - name: branding
+ configMap:
+ name: switchboard-branding
+ optional: true
---
apiVersion: v1
kind: Service
diff --git a/nginx.frontend.conf b/nginx.frontend.conf
deleted file mode 100644
index 813757c..0000000
--- a/nginx.frontend.conf
+++ /dev/null
@@ -1,29 +0,0 @@
-server {
- listen 80;
- server_name _;
- root /usr/share/nginx/html;
- index index.html;
-
- # Gzip compression
- gzip on;
- gzip_vary on;
- gzip_min_length 1024;
- gzip_proxied expired no-cache no-store private auth;
- gzip_types text/plain text/css text/xml text/javascript application/x-javascript application/xml application/javascript;
-
- # Cache static assets
- location ~* \.(jpg|jpeg|png|gif|ico|css|js|woff|woff2)$ {
- expires 1y;
- add_header Cache-Control "public, immutable";
- }
-
- # SPA fallback
- location / {
- try_files $uri $uri/ /index.html;
- }
-
- # Security headers
- add_header X-Frame-Options "SAMEORIGIN" always;
- add_header X-Content-Type-Options "nosniff" always;
- add_header X-XSS-Protection "1; mode=block" always;
-}
diff --git a/server/database/migrations/015_preset_avatars.sql b/server/database/migrations/015_preset_avatars.sql
new file mode 100644
index 0000000..be40fb1
--- /dev/null
+++ b/server/database/migrations/015_preset_avatars.sql
@@ -0,0 +1,6 @@
+-- Avatar support for model presets.
+-- Users table already has avatar_url from 001_full_schema.sql.
+
+ALTER TABLE model_presets ADD COLUMN IF NOT EXISTS avatar TEXT DEFAULT '';
+
+COMMENT ON COLUMN model_presets.avatar IS 'Base64 data URI of preset avatar image (128x128 PNG), empty = use icon emoji';
diff --git a/server/handlers/admin.go b/server/handlers/admin.go
index 92ef6e9..9be9d76 100644
--- a/server/handlers/admin.go
+++ b/server/handlers/admin.go
@@ -283,6 +283,7 @@ func (h *AdminHandler) DeleteUser(c *gin.Context) {
var publicSettingKeys = map[string]bool{
"banner": true,
+ "branding": true,
"user_providers_enabled": true,
"registration_enabled": true,
"registration_default_state": true,
diff --git a/server/handlers/apiconfigs.go b/server/handlers/apiconfigs.go
index 4e0bc24..8513698 100644
--- a/server/handlers/apiconfigs.go
+++ b/server/handlers/apiconfigs.go
@@ -435,6 +435,7 @@ func (h *APIConfigHandler) ListEnabledModels(c *gin.Context) {
IsPreset bool `json:"is_preset,omitempty"`
PresetID string `json:"preset_id,omitempty"`
PresetScope string `json:"preset_scope,omitempty"`
+ PresetAvatar string `json:"preset_avatar,omitempty"`
}
models := make([]enabledModel, 0)
@@ -538,7 +539,7 @@ func (h *APIConfigHandler) ListEnabledModels(c *gin.Context) {
// ── 3. Active presets (global + user's personal + shared) ──
presetRows, err := database.DB.Query(`
SELECT mp.id, mp.name, mp.description, mp.base_model_id, mp.api_config_id,
- mp.icon, mp.scope, mp.temperature, mp.max_tokens,
+ mp.icon, mp.avatar, mp.scope, mp.temperature, mp.max_tokens,
COALESCE(ac.provider, '') as provider, COALESCE(ac.name, '') as provider_name
FROM model_presets mp
LEFT JOIN api_configs ac ON mp.api_config_id = ac.id
@@ -553,12 +554,12 @@ func (h *APIConfigHandler) ListEnabledModels(c *gin.Context) {
if err == nil {
defer presetRows.Close()
for presetRows.Next() {
- var presetID, name, description, baseModelID, icon, scope, provID, provName string
+ var presetID, name, description, baseModelID, icon, avatar, scope, provID, provName string
var apiConfigID *string
var temp *float64
var maxTok *int
if err := presetRows.Scan(&presetID, &name, &description, &baseModelID, &apiConfigID,
- &icon, &scope, &temp, &maxTok, &provID, &provName); err != nil {
+ &icon, &avatar, &scope, &temp, &maxTok, &provID, &provName); err != nil {
continue
}
@@ -615,6 +616,7 @@ func (h *APIConfigHandler) ListEnabledModels(c *gin.Context) {
IsPreset: true,
PresetID: presetID,
PresetScope: scope,
+ PresetAvatar: avatar,
})
}
}
diff --git a/server/handlers/auth.go b/server/handlers/auth.go
index a159e19..e964a03 100644
--- a/server/handlers/auth.go
+++ b/server/handlers/auth.go
@@ -64,6 +64,7 @@ type userResponse struct {
Email string `json:"email"`
DisplayName *string `json:"display_name"`
Role string `json:"role"`
+ Avatar *string `json:"avatar,omitempty"`
}
// AuthHandler holds dependencies for auth endpoints.
@@ -129,9 +130,9 @@ func (h *AuthHandler) Register(c *gin.Context) {
err = database.DB.QueryRow(`
INSERT INTO users (username, email, password_hash, role, is_active)
VALUES ($1, $2, $3, $4, $5)
- RETURNING id, username, email, display_name, role
+ RETURNING id, username, email, display_name, role, avatar_url
`, req.Username, req.Email, string(hash), role, isActive).Scan(
- &user.ID, &user.Username, &user.Email, &user.DisplayName, &user.Role,
+ &user.ID, &user.Username, &user.Email, &user.DisplayName, &user.Role, &user.Avatar,
)
if err != nil {
if strings.Contains(err.Error(), "duplicate key") {
@@ -268,12 +269,12 @@ func (h *AuthHandler) Login(c *gin.Context) {
var isActive bool
err := database.DB.QueryRow(`
- SELECT id, username, email, display_name, role, password_hash, is_active
+ SELECT id, username, email, display_name, role, avatar_url, password_hash, is_active
FROM users
WHERE email = $1 OR username = $1
`, req.Login).Scan(
&user.ID, &user.Username, &user.Email, &user.DisplayName,
- &user.Role, &passwordHash, &isActive,
+ &user.Role, &user.Avatar, &passwordHash, &isActive,
)
if err == sql.ErrNoRows {
c.JSON(http.StatusUnauthorized, gin.H{"error": "invalid credentials"})
@@ -352,10 +353,10 @@ func (h *AuthHandler) Refresh(c *gin.Context) {
var user userResponse
var isActive bool
err = database.DB.QueryRow(`
- SELECT id, username, email, display_name, role, is_active
+ SELECT id, username, email, display_name, role, avatar_url, is_active
FROM users WHERE id = $1
`, userID).Scan(
- &user.ID, &user.Username, &user.Email, &user.DisplayName, &user.Role, &isActive,
+ &user.ID, &user.Username, &user.Email, &user.DisplayName, &user.Role, &user.Avatar, &isActive,
)
if err != nil || !isActive {
c.JSON(http.StatusUnauthorized, gin.H{"error": "account unavailable"})
diff --git a/server/handlers/avatar.go b/server/handlers/avatar.go
new file mode 100644
index 0000000..0522cbf
--- /dev/null
+++ b/server/handlers/avatar.go
@@ -0,0 +1,245 @@
+package handlers
+
+import (
+ "bytes"
+ "encoding/base64"
+ "image"
+ "image/color"
+ "image/png"
+ "net/http"
+ "strings"
+
+ // Register decoders so image.Decode works for common formats
+ _ "image/gif"
+ _ "image/jpeg"
+
+ "github.com/gin-gonic/gin"
+
+ "git.gobha.me/xcaliber/chat-switchboard/database"
+)
+
+const avatarSize = 128
+const maxUploadBytes = 2 * 1024 * 1024 // 2MB raw input limit
+
+type uploadAvatarRequest struct {
+ Image string `json:"image" binding:"required"` // base64 data URI or raw base64
+}
+
+// ── Upload Avatar ───────────────────────────
+// POST /api/v1/profile/avatar
+// Accepts { "image": "data:image/png;base64,..." } or { "image": "<raw base64>" }
+// Decodes, resizes to 128×128 PNG, stores as data URI.
+func (h *SettingsHandler) UploadAvatar(c *gin.Context) {
+ userID := getUserID(c)
+
+ var req uploadAvatarRequest
+ if err := c.ShouldBindJSON(&req); err != nil {
+ c.JSON(http.StatusBadRequest, gin.H{"error": "missing image field"})
+ return
+ }
+
+ // Strip data URI prefix if present
+ b64 := req.Image
+ if idx := strings.Index(b64, ","); idx >= 0 {
+ b64 = b64[idx+1:]
+ }
+
+ // Decode base64
+ raw, err := base64.StdEncoding.DecodeString(b64)
+ if err != nil {
+ c.JSON(http.StatusBadRequest, gin.H{"error": "invalid base64 encoding"})
+ return
+ }
+ if len(raw) > maxUploadBytes {
+ c.JSON(http.StatusBadRequest, gin.H{"error": "image too large (max 2MB)"})
+ return
+ }
+
+ // Decode image (supports PNG, JPEG, GIF via registered decoders)
+ src, _, err := image.Decode(bytes.NewReader(raw))
+ if err != nil {
+ c.JSON(http.StatusBadRequest, gin.H{"error": "unsupported image format"})
+ return
+ }
+
+ // Resize to 128×128
+ resized := resizeBilinear(src, avatarSize, avatarSize)
+
+ // Encode to PNG
+ var buf bytes.Buffer
+ if err := png.Encode(&buf, resized); err != nil {
+ c.JSON(http.StatusInternalServerError, gin.H{"error": "failed to encode avatar"})
+ return
+ }
+
+ // Build data URI
+ dataURI := "data:image/png;base64," + base64.StdEncoding.EncodeToString(buf.Bytes())
+
+ // Store in DB
+ _, err = database.DB.Exec(
+ `UPDATE users SET avatar_url = $1, updated_at = NOW() WHERE id = $2`,
+ dataURI, userID,
+ )
+ if err != nil {
+ c.JSON(http.StatusInternalServerError, gin.H{"error": "failed to save avatar"})
+ return
+ }
+
+ c.JSON(http.StatusOK, gin.H{"avatar": dataURI})
+}
+
+// ── Delete Avatar ───────────────────────────
+// DELETE /api/v1/profile/avatar
+func (h *SettingsHandler) DeleteAvatar(c *gin.Context) {
+ userID := getUserID(c)
+
+ _, err := database.DB.Exec(
+ `UPDATE users SET avatar_url = NULL, updated_at = NOW() WHERE id = $1`,
+ userID,
+ )
+ if err != nil {
+ c.JSON(http.StatusInternalServerError, gin.H{"error": "failed to remove avatar"})
+ return
+ }
+
+ c.JSON(http.StatusOK, gin.H{"message": "avatar removed"})
+}
+
+// ── Bilinear Resize ─────────────────────────
+// Pure stdlib bilinear interpolation. Quality is fine for 128×128 avatars.
+func resizeBilinear(src image.Image, w, h int) *image.RGBA {
+ dst := image.NewRGBA(image.Rect(0, 0, w, h))
+ sb := src.Bounds()
+ sw := float64(sb.Dx())
+ sh := float64(sb.Dy())
+
+ for y := 0; y < h; y++ {
+ for x := 0; x < w; x++ {
+ // Map destination pixel to source coordinates
+ sx := (float64(x) + 0.5) * sw / float64(w) - 0.5
+ sy := (float64(y) + 0.5) * sh / float64(h) - 0.5
+
+ x0 := int(sx)
+ y0 := int(sy)
+ xf := sx - float64(x0)
+ yf := sy - float64(y0)
+
+ // Clamp
+ if x0 < sb.Min.X { x0 = sb.Min.X }
+ if y0 < sb.Min.Y { y0 = sb.Min.Y }
+ x1 := x0 + 1
+ y1 := y0 + 1
+ if x1 >= sb.Max.X { x1 = sb.Max.X - 1 }
+ if y1 >= sb.Max.Y { y1 = sb.Max.Y - 1 }
+
+ // Sample 4 neighbors
+ c00 := src.At(x0, y0)
+ c10 := src.At(x1, y0)
+ c01 := src.At(x0, y1)
+ c11 := src.At(x1, y1)
+
+ dst.Set(x, y, bilinearMix(c00, c10, c01, c11, xf, yf))
+ }
+ }
+ return dst
+}
+
+func bilinearMix(c00, c10, c01, c11 color.Color, xf, yf float64) color.Color {
+ r00, g00, b00, a00 := c00.RGBA()
+ r10, g10, b10, a10 := c10.RGBA()
+ r01, g01, b01, a01 := c01.RGBA()
+ r11, g11, b11, a11 := c11.RGBA()
+
+ mix := func(v00, v10, v01, v11 uint32) uint8 {
+ top := float64(v00)*(1-xf) + float64(v10)*xf
+ bot := float64(v01)*(1-xf) + float64(v11)*xf
+ return uint8((top*(1-yf) + bot*yf) / 256)
+ }
+
+ return color.RGBA{
+ R: mix(r00, r10, r01, r11),
+ G: mix(g00, g10, g01, g11),
+ B: mix(b00, b10, b01, b11),
+ A: mix(a00, a10, a01, a11),
+ }
+}
+
+// ── Preset Avatar Upload ────────────────────
+// POST /api/v1/presets/:id/avatar (user) or /api/v1/admin/presets/:id/avatar (admin)
+func UploadPresetAvatar(c *gin.Context) {
+ presetID := c.Param("id")
+
+ var req uploadAvatarRequest
+ if err := c.ShouldBindJSON(&req); err != nil {
+ c.JSON(http.StatusBadRequest, gin.H{"error": "missing image field"})
+ return
+ }
+
+ b64 := req.Image
+ if idx := strings.Index(b64, ","); idx >= 0 {
+ b64 = b64[idx+1:]
+ }
+
+ raw, err := base64.StdEncoding.DecodeString(b64)
+ if err != nil {
+ c.JSON(http.StatusBadRequest, gin.H{"error": "invalid base64 encoding"})
+ return
+ }
+ if len(raw) > maxUploadBytes {
+ c.JSON(http.StatusBadRequest, gin.H{"error": "image too large (max 2MB)"})
+ return
+ }
+
+ src, _, err := image.Decode(bytes.NewReader(raw))
+ if err != nil {
+ c.JSON(http.StatusBadRequest, gin.H{"error": "unsupported image format"})
+ return
+ }
+
+ resized := resizeBilinear(src, avatarSize, avatarSize)
+
+ var buf bytes.Buffer
+ if err := png.Encode(&buf, resized); err != nil {
+ c.JSON(http.StatusInternalServerError, gin.H{"error": "failed to encode avatar"})
+ return
+ }
+
+ dataURI := "data:image/png;base64," + base64.StdEncoding.EncodeToString(buf.Bytes())
+
+ result, err := database.DB.Exec(
+ `UPDATE model_presets SET avatar = $1, updated_at = NOW() WHERE id = $2`,
+ dataURI, presetID,
+ )
+ if err != nil {
+ c.JSON(http.StatusInternalServerError, gin.H{"error": "failed to save avatar"})
+ return
+ }
+ rows, _ := result.RowsAffected()
+ if rows == 0 {
+ c.JSON(http.StatusNotFound, gin.H{"error": "preset not found"})
+ return
+ }
+
+ c.JSON(http.StatusOK, gin.H{"avatar": dataURI})
+}
+
+// DeletePresetAvatar clears a preset's avatar.
+func DeletePresetAvatar(c *gin.Context) {
+ presetID := c.Param("id")
+
+ result, err := database.DB.Exec(
+ `UPDATE model_presets SET avatar = '', updated_at = NOW() WHERE id = $1`,
+ presetID,
+ )
+ if err != nil {
+ c.JSON(http.StatusInternalServerError, gin.H{"error": "failed to remove avatar"})
+ return
+ }
+ rows, _ := result.RowsAffected()
+ if rows == 0 {
+ c.JSON(http.StatusNotFound, gin.H{"error": "preset not found"})
+ return
+ }
+
+ c.JSON(http.StatusOK, gin.H{"message": "avatar removed"})
+}
diff --git a/server/handlers/presets.go b/server/handlers/presets.go
index ca713c1..52da00d 100644
--- a/server/handlers/presets.go
+++ b/server/handlers/presets.go
@@ -67,7 +67,7 @@ func (h *PresetHandler) ListUserPresets(c *gin.Context) {
SELECT mp.id, mp.name, mp.description, mp.base_model_id, mp.api_config_id,
mp.system_prompt, mp.temperature, mp.max_tokens, mp.tools_enabled,
mp.scope, mp.team_id, mp.created_by, mp.is_shared, mp.is_active,
- mp.icon, mp.created_at, mp.updated_at,
+ mp.icon, mp.avatar, mp.created_at, mp.updated_at,
COALESCE(ac.name, '') as provider_name
FROM model_presets mp
LEFT JOIN api_configs ac ON mp.api_config_id = ac.id
@@ -92,7 +92,7 @@ func (h *PresetHandler) ListUserPresets(c *gin.Context) {
&p.ID, &p.Name, &p.Description, &p.BaseModelID, &p.APIConfigID,
&p.SystemPrompt, &p.Temperature, &p.MaxTokens, &p.ToolsEnabled,
&p.Scope, &p.TeamID, &p.CreatedBy, &p.IsShared, &p.IsActive,
- &p.Icon, &p.CreatedAt, &p.UpdatedAt, &p.ProviderName,
+ &p.Icon, &p.Avatar, &p.CreatedAt, &p.UpdatedAt, &p.ProviderName,
); err != nil {
continue
}
@@ -281,7 +281,7 @@ func (h *PresetHandler) ListAdminPresets(c *gin.Context) {
SELECT mp.id, mp.name, mp.description, mp.base_model_id, mp.api_config_id,
mp.system_prompt, mp.temperature, mp.max_tokens, mp.tools_enabled,
mp.scope, mp.team_id, mp.created_by, mp.is_shared, mp.is_active,
- mp.icon, mp.created_at, mp.updated_at,
+ mp.icon, mp.avatar, mp.created_at, mp.updated_at,
COALESCE(ac.name, '') as provider_name,
COALESCE(u.username, '') as creator_name
FROM model_presets mp
@@ -307,7 +307,7 @@ func (h *PresetHandler) ListAdminPresets(c *gin.Context) {
&p.ID, &p.Name, &p.Description, &p.BaseModelID, &p.APIConfigID,
&p.SystemPrompt, &p.Temperature, &p.MaxTokens, &p.ToolsEnabled,
&p.Scope, &p.TeamID, &p.CreatedBy, &p.IsShared, &p.IsActive,
- &p.Icon, &p.CreatedAt, &p.UpdatedAt, &p.ProviderName, &p.CreatorName,
+ &p.Icon, &p.Avatar, &p.CreatedAt, &p.UpdatedAt, &p.ProviderName, &p.CreatorName,
); err != nil {
continue
}
diff --git a/server/handlers/settings.go b/server/handlers/settings.go
index 5f01b34..cce0502 100644
--- a/server/handlers/settings.go
+++ b/server/handlers/settings.go
@@ -30,6 +30,7 @@ type profileResponse struct {
Email string `json:"email"`
DisplayName *string `json:"display_name"`
Role string `json:"role"`
+ Avatar *string `json:"avatar,omitempty"`
Settings map[string]interface{} `json:"settings"`
CreatedAt string `json:"created_at"`
}
@@ -50,11 +51,11 @@ func (h *SettingsHandler) GetProfile(c *gin.Context) {
var p profileResponse
var settingsRaw string
err := database.DB.QueryRow(`
- SELECT id, username, email, display_name, role, settings::text, created_at
+ SELECT id, username, email, display_name, role, avatar_url, settings::text, created_at
FROM users WHERE id = $1
`, userID).Scan(
&p.ID, &p.Username, &p.Email, &p.DisplayName, &p.Role,
- &settingsRaw, &p.CreatedAt,
+ &p.Avatar, &settingsRaw, &p.CreatedAt,
)
if err == sql.ErrNoRows {
c.JSON(http.StatusNotFound, gin.H{"error": "user not found"})
diff --git a/server/main.go b/server/main.go
index 842015d..5ad54c5 100644
--- a/server/main.go
+++ b/server/main.go
@@ -134,6 +134,8 @@ func main() {
protected.GET("/profile", settings.GetProfile)
protected.PUT("/profile", settings.UpdateProfile)
protected.POST("/profile/password", settings.ChangePassword)
+ protected.POST("/profile/avatar", settings.UploadAvatar)
+ protected.DELETE("/profile/avatar", settings.DeleteAvatar)
protected.GET("/settings", settings.GetSettings)
protected.PUT("/settings", settings.UpdateSettings)
@@ -143,6 +145,8 @@ func main() {
protected.POST("/presets", presets.CreateUserPreset)
protected.PUT("/presets/:id", presets.UpdateUserPreset)
protected.DELETE("/presets/:id", presets.DeleteUserPreset)
+ protected.POST("/presets/:id/avatar", handlers.UploadPresetAvatar)
+ protected.DELETE("/presets/:id/avatar", handlers.DeletePresetAvatar)
// Notes
notes := handlers.NewNoteHandler()
@@ -202,6 +206,8 @@ func main() {
admin.POST("/presets", presetAdm.CreateAdminPreset)
admin.PUT("/presets/:id", presetAdm.UpdateAdminPreset)
admin.DELETE("/presets/:id", presetAdm.DeleteAdminPreset)
+ admin.POST("/presets/:id/avatar", handlers.UploadPresetAvatar)
+ admin.DELETE("/presets/:id/avatar", handlers.DeletePresetAvatar)
}
}
diff --git a/server/models/models.go b/server/models/models.go
index 1dfea8e..0dd8479 100644
--- a/server/models/models.go
+++ b/server/models/models.go
@@ -174,6 +174,7 @@ type ModelPreset struct {
IsShared bool `json:"is_shared" db:"is_shared"`
IsActive bool `json:"is_active" db:"is_active"`
Icon string `json:"icon,omitempty" db:"icon"`
+ Avatar string `json:"avatar,omitempty" db:"avatar"`
}
// ── Settings ────────────────────────────────
diff --git a/src/css/styles.css b/src/css/styles.css
index 26d3754..10cc991 100644
--- a/src/css/styles.css
+++ b/src/css/styles.css
@@ -276,6 +276,10 @@ a:hover { text-decoration: underline; }
background: var(--bg-raised); display: flex; align-items: center;
justify-content: center; font-size: 13px; font-weight: 600;
color: var(--accent); flex-shrink: 0; position: relative;
+ overflow: hidden;
+}
+.user-avatar-img {
+ width: 100%; height: 100%; object-fit: cover; border-radius: 50%;
}
.avatar-bug {
position: absolute; bottom: -3px; right: -3px;
@@ -369,6 +373,7 @@ a:hover { text-decoration: underline; }
.model-dropdown-item.selected { background: var(--bg-hover); color: var(--accent); }
.model-dropdown-item .item-label { flex: 1; overflow: hidden; text-overflow: ellipsis; }
.model-dropdown-item .item-provider { margin-left: auto; font-size: 10px; color: var(--text-3); }
+.dropdown-avatar { width: 18px; height: 18px; border-radius: 50%; object-fit: cover; flex-shrink: 0; }
.model-caps {
display: flex; align-items: center; gap: 4px;
@@ -408,6 +413,10 @@ a:hover { text-decoration: underline; }
width: 28px; height: 28px; border-radius: 50%;
display: flex; align-items: center; justify-content: center;
font-size: 14px; flex-shrink: 0; margin-top: 2px;
+ overflow: hidden;
+}
+.msg-avatar-img {
+ width: 100%; height: 100%; object-fit: cover; border-radius: 50%;
}
.message.user .msg-avatar { background: var(--accent-dim); }
.message.assistant .msg-avatar { background: var(--bg-raised); }
@@ -725,8 +734,10 @@ button { font-family: var(--font); cursor: pointer; }
.hero-content { position: relative; z-index: 1; max-width: 520px; }
.hero-logo-row { display: flex; align-items: center; gap: 14px; margin-bottom: 1.5rem; }
.hero-logo-mark { width: 48px; height: 48px; flex-shrink: 0; }
+.hero-logo-img { width: 48px; height: 48px; object-fit: contain; border-radius: 8px; }
.hero-wordmark { font-size: 1.6rem; font-weight: 700; letter-spacing: -0.03em; line-height: 1.1; }
.hero-wordmark span { color: var(--accent); }
+.brand-logo-img { width: 18px; height: 18px; object-fit: contain; border-radius: 3px; vertical-align: middle; }
.hero-headline {
font-size: 2.4rem; font-weight: 700; line-height: 1.15;
letter-spacing: -0.03em; margin-bottom: 1rem;
@@ -808,6 +819,7 @@ button { font-family: var(--font); cursor: pointer; }
.splash-hero { padding: 1.5rem 1.25rem 1rem; }
.hero-logo-row { gap: 10px; margin-bottom: 0.75rem; }
.hero-logo-mark { width: 40px; height: 40px; }
+ .hero-logo-img { width: 40px; height: 40px; }
.hero-wordmark { font-size: 1.3rem; }
.hero-headline { font-size: 1.35rem; margin-bottom: 0.5rem; }
.hero-sub { font-size: 0.85rem; margin-bottom: 1rem; line-height: 1.5; }
@@ -891,6 +903,17 @@ button { font-family: var(--font); cursor: pointer; }
.settings-section { margin-bottom: 1.5rem; padding-bottom: 1rem; border-bottom: 1px solid var(--border); }
.settings-section:last-child { border-bottom: none; margin-bottom: 0; }
.settings-section h3 { font-size: 12px; font-weight: 600; color: var(--text-3); margin-bottom: 12px; text-transform: uppercase; letter-spacing: 0.06em; }
+.avatar-upload-row { display: flex; align-items: center; gap: 14px; margin-bottom: 1rem; }
+.avatar-preview {
+ width: 56px; height: 56px; border-radius: 50%; background: var(--bg-raised);
+ display: flex; align-items: center; justify-content: center;
+ font-size: 22px; font-weight: 600; color: var(--accent); flex-shrink: 0;
+ overflow: hidden; border: 2px solid var(--border);
+}
+.avatar-preview img { width: 100%; height: 100%; object-fit: cover; }
+.avatar-actions { display: flex; flex-direction: column; gap: 4px; }
+.avatar-actions .form-hint { font-size: 11px; margin-left: 0; }
+.avatar-preview-sm { width: 42px; height: 42px; font-size: 18px; }
/* Modal tab bars — shared horizontal scroll with arrow navigation */
.modal-tabs-wrap {
@@ -1001,6 +1024,8 @@ button { font-family: var(--font); cursor: pointer; }
.admin-preset-row { display: flex; align-items: center; gap: 12px; padding: 10px 0; border-bottom: 1px solid var(--border); font-size: 14px; }
.admin-preset-row .preset-info { flex: 1; min-width: 0; }
+.preset-row-avatar { width: 22px; height: 22px; border-radius: 50%; object-fit: cover; vertical-align: middle; margin-right: 6px; }
+.preset-row-icon { margin-right: 4px; }
.admin-preset-row .preset-meta { font-size: 12px; color: var(--text-3); margin-top: 2px; }
.admin-preset-row .preset-desc { font-size: 12px; color: var(--text-2); margin-top: 2px; overflow: hidden; text-overflow: ellipsis; white-space: nowrap; }
.admin-preset-row .preset-actions { display: flex; gap: 6px; align-items: center; flex-shrink: 0; }
diff --git a/src/index.html b/src/index.html
index 6e586e1..8000938 100644
--- a/src/index.html
+++ b/src/index.html
@@ -6,11 +6,13 @@
<base href="%%BASE_HREF%%">
<script>window.__BASE__ = '%%BASE_PATH%%';</script>
<script>window.__VERSION__ = '%%APP_VERSION%%'; if (window.__VERSION__.includes('%%')) window.__VERSION__ = 'dev';</script>
+ <script>try { window.__BRANDING__ = %%BRANDING_JSON%%; } catch(e) { window.__BRANDING__ = {}; }</script>
<title>Chat Switchboard
+
@@ -25,9 +27,9 @@