Rc/v5.14.0

CHANGELOG.md

@@ -7,6 +7,16 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [5.14.0] - 2026-09-09
+
+### Added
+
+- **OrcaRouter AI provider** — Added `orcarouter` as a named AI provider option (OpenAI-compatible endpoint, standard Bearer-token auth). Admins can select it in System Settings → AI Helper or set `AI_PROVIDER=orcarouter` with `AI_BASE_URL=https://api.orcarouter.ai`.
+
+### Documentation
+
+- **Version** — Bumped `setup.py` to **5.14.0** (single source of truth for the application version).
+
 ## [5.13.5] - 2026-09-08
 
 ### Fixed

README.md

@@ -105,6 +105,10 @@ TimeTracker has been continuously enhanced with powerful new features! Here's wh
 
 **Current version** is defined in `setup.py` (single source of truth). See [CHANGELOG.md](CHANGELOG.md) for versioned release history.
 
+### ✨ Highlights of v5.14.0
+
+**Minor (5.14.0):** **OrcaRouter AI provider** — `orcarouter` is now a named AI provider option (OpenAI-compatible endpoint, Bearer-token auth). Select it in System Settings → AI Helper or set `AI_PROVIDER=orcarouter` with `AI_BASE_URL=https://api.orcarouter.ai`. See [CHANGELOG.md](CHANGELOG.md#5140---2026-09-09).
+
 ### ✨ Highlights of v5.13.5
 
 **Patch (5.13.5):** **Timezone `local_date`** — date-only values (e.g. task due dates) no longer crash the dashboard filter. **Idle skip paused (#752)** — paused timers are excluded from idle notify / needs-review / auto-stop. **Manual entry single-client (#753)** — restores the readonly pre-filled client lock when the org has one client. See [CHANGELOG.md](CHANGELOG.md#5135---2026-09-08).

app/models/settings.py

@@ -467,7 +467,7 @@ def cfg(name, default=None):
 
         return {
             "enabled": bool(enabled),
-            "provider": provider if provider in {"ollama", "openai_compatible"} else "ollama",
+            "provider": provider if provider in {"ollama", "openai_compatible", "orcarouter"} else "ollama",
             "base_url": base_url.rstrip("/"),
             "model": model,
             "api_key": api_key if include_secrets else "",

app/routes/admin.py

@@ -1645,7 +1645,7 @@ def settings():
                 settings_obj.ai_enabled = None
 
             ai_provider = (request.form.get("ai_provider") or "ollama").strip().lower()
-            if ai_provider not in ("ollama", "openai_compatible"):
+            if ai_provider not in ("ollama", "openai_compatible", "orcarouter"):
                 ai_provider = "ollama"
             settings_obj.ai_provider = ai_provider
             settings_obj.ai_base_url = (request.form.get("ai_base_url") or "").strip()

app/services/llm_service.py

@@ -80,7 +80,7 @@ def ensure_enabled(self) -> None:
             raise AIServiceError("AI helper is disabled.", "ai_disabled", 503)
         if not self.config.base_url or not self.config.model:
             raise AIServiceError("AI helper is not fully configured.", "ai_not_configured", 400)
-        if self.config.provider == "openai_compatible" and not self.config.api_key:
+        if self.config.provider in ("openai_compatible", "orcarouter") and not self.config.api_key:
             raise AIServiceError("Hosted AI provider requires an API key.", "ai_missing_api_key", 400)
 
     def test_connection(self) -> Dict[str, Any]:
@@ -186,7 +186,7 @@ def confirm_action(self, user: User, action: Dict[str, Any]) -> Dict[str, Any]:
     def _chat_completion(self, messages: List[Dict[str, str]], max_tokens: int = 700) -> Dict[str, Any]:
         url = f"{self.config.base_url.rstrip('/')}/v1/chat/completions"
         headers = {"Accept": "application/json", "Content-Type": "application/json"}
-        if self.config.provider == "openai_compatible" and self.config.api_key:
+        if self.config.provider in ("openai_compatible", "orcarouter") and self.config.api_key:
             headers["Authorization"] = f"Bearer {self.config.api_key}"
         payload = {
             "model": self.config.model,

app/templates/admin/settings.html

@@ -708,12 +708,13 @@ <h2 class="text-lg font-semibold mb-4">{{ _('AI Helper') }}</h2>
                         <select name="ai_provider" id="ai_provider" class="form-input">
                             <option value="ollama" {% if ai_config.provider == 'ollama' %}selected{% endif %}>Ollama / local OpenAI-compatible</option>
                             <option value="openai_compatible" {% if ai_config.provider == 'openai_compatible' %}selected{% endif %}>Hosted OpenAI-compatible</option>
+                            <option value="orcarouter" {% if ai_config.provider == 'orcarouter' %}selected{% endif %}>OrcaRouter (hosted, OpenAI-compatible)</option>
                         </select>
                     </div>
                     <div>
                         <label for="ai_base_url" class="form-label">{{ _('Base URL') }}</label>
                         <input type="url" name="ai_base_url" id="ai_base_url" value="{{ settings.ai_base_url or ai_config.base_url }}" placeholder="http://127.0.0.1:11434" class="form-input">
-                        <p class="mt-1 text-xs text-text-muted-light dark:text-text-muted-dark">{{ _('For Ollama, this is the URL from the Flask server to Ollama.') }}</p>
+                        <p class="mt-1 text-xs text-text-muted-light dark:text-text-muted-dark">{{ _('For Ollama, this is the URL from the Flask server to Ollama. For OrcaRouter, use https://api.orcarouter.ai.') }}</p>
                     </div>
                     <div>
                         <label for="ai_model" class="form-label">{{ _('Model') }}</label>

docs/mobile-desktop-apps/DESKTOP_WEBAPP_GAP.md

@@ -6,9 +6,9 @@ Living checklist for bringing the Electron desktop client up to speed with the F
 
 | Component | Version | Source |
 |-----------|---------|--------|
-| Web / backend | 5.13.5 | `setup.py` |
-| Desktop | 5.13.5 | `desktop/package.json` |
-| Mobile | 5.13.5 | `mobile/pubspec.yaml` |
+| Web / backend | 5.14.0 | `setup.py` |
+| Desktop | 5.14.0 | `desktop/package.json` |
+| Mobile | 5.14.0 | `mobile/pubspec.yaml` |
 
 ## What desktop has
 

env.example

@@ -73,6 +73,13 @@ AI_ENABLED=false
 # AI_MODEL=...
 # AI_API_KEY=...
 
+# --- Optional: OrcaRouter (hosted, OpenAI-compatible; failover + budgets) ---
+# AI_ENABLED=true
+# AI_PROVIDER=orcarouter
+# AI_BASE_URL=https://api.orcarouter.ai
+# AI_MODEL=<model-slug>
+# AI_API_KEY=<your-orcarouter-api-key>
+
 AI_PROVIDER=ollama
 AI_BASE_URL=http://ollama:11434
 AI_MODEL=llama3.1

setup.py

@@ -7,7 +7,7 @@
 
 setup(
     name='timetracker',
-    version='5.13.5',
+    version='5.14.0',
     packages=find_packages(),
     include_package_data=True,
     package_data={