chore: release v1.7.2 — Gemini model import fix, Pino transport fallback

CHANGELOG.md

@@ -7,6 +7,15 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ---
 
+## [1.7.2] — 2026-03-01
+
+### 🐛 Fixed
+
+- **Gemini Imported Models Return 404** — Gemini API returns model IDs with `models/` prefix (e.g. `models/gemini-2.5-pro`), but the execution base URL already ends with `/v1beta/models`. This caused doubles paths like `.../v1beta/models/models/gemini-...` → HTTP 404. Fixed by stripping the `models/` prefix during import in `parseResponse` for both `gemini` and `gemini-cli` providers ([#163](https://github.com/diegosouzapw/OmniRoute/issues/163))
+- **Pino File Transport Fails in Next.js Production** — Console Logs dashboard page returned HTTP 500 because `pino.transport()` (worker-thread-based) fails silently inside Next.js production bundles. The catch block now logs the actual error message for diagnostics, and includes a sync `pino.destination()` + `pino.multistream()` fallback that works in bundled environments ([#165](https://github.com/diegosouzapw/OmniRoute/issues/165))
+
+---
+
 ## [1.7.1] — 2026-02-28
 
 ### 🐛 Fixed

package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "omniroute",
-  "version": "1.7.0",
+  "version": "1.7.2",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "omniroute",
-      "version": "1.7.0",
+      "version": "1.7.2",
       "hasInstallScript": true,
       "license": "MIT",
       "workspaces": [

package.json

@@ -1,6 +1,6 @@
 {
   "name": "omniroute",
-  "version": "1.7.1",
+  "version": "1.7.2",
   "description": "Smart AI Router with auto fallback — route to FREE & cheap models, zero downtime. Works with Cursor, Cline, Claude Desktop, Codex, and any OpenAI-compatible tool.",
   "type": "module",
   "bin": {