chore: Bump version to 0.3.2

Update version references across all project files for v0.3.2 release.

Files updated:
- VERSION
- pyproject.toml
- frontend/package.json
- frontend/package-lock.json
- docs-site/blog/2025-12-17-v0.3.2-release.md (new)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Author: davidamacey

VERSION

@@ -1 +1 @@
-v0.3.1
+v0.3.2

docs-site/blog/2025-12-17-v0.3.2-release.md

@@ -0,0 +1,53 @@
+---
+slug: v0.3.2-release
+title: "v0.3.2 - Setup Script Bug Fixes"
+authors: [davidamacey]
+tags: [release, bugfix, setup]
+---
+
+# OpenTranscribe v0.3.2 Released
+
+Patch release fixing critical bugs in the one-liner installation script that prevented successful setup on fresh installations.
+
+<!-- truncate -->
+
+## Overview
+
+This is a **scripts-only release** - no Docker container rebuild required. Existing installations are unaffected as the containers already have the necessary fixes.
+
+## Fixed
+
+### Scripts Directory Creation
+
+Fixed `curl` error 23 ("Failure writing output to destination") when downloading SSL and permission scripts during setup. The `scripts/` directory is now created before attempting downloads.
+
+**Before:** Setup would fail with cryptic curl errors
+**After:** All script downloads complete successfully
+
+### PyTorch 2.6+ Compatibility
+
+Applied the `torch.load` patch to `download-models.py` for PyTorch 2.6+ compatibility. This mirrors the fix already present in the backend (from Wes Brown's commit).
+
+PyTorch 2.6 changed the default value of `weights_only` from `False` to `True`, which caused omegaconf deserialization errors when downloading PyAnnote speaker diarization models.
+
+**Before:** Model downloads failed with "WeightsUnpickler error: Unsupported global: GLOBAL omegaconf.listconfig.ListConfig"
+**After:** Models download successfully
+
+## Upgrade Notes
+
+### New Installations
+
+The one-liner setup script now works correctly:
+
+```bash
+curl -fsSL https://raw.githubusercontent.com/davidamacey/OpenTranscribe/master/setup-opentranscribe.sh | bash
+```
+
+### Existing Installations
+
+No action required - Docker containers already have the PyTorch fix in the backend.
+
+## Links
+
+- [Full Changelog](https://github.com/davidamacey/OpenTranscribe/blob/master/CHANGELOG.md)
+- [GitHub Release](https://github.com/davidamacey/OpenTranscribe/releases/tag/v0.3.2)

frontend/package-lock.json

@@ -1,7 +1,7 @@
 {
   "name": "opentranscribe-frontend",
   "private": true,
-  "version": "0.3.1",
+  "version": "0.3.2",
   "type": "module",
   "scripts": {
     "dev": "vite dev",

frontend/package.json

@@ -2,7 +2,7 @@
 
 [project]
 name = "opentranscribe"
-version = "0.3.1"
+version = "0.3.2"
 description = "AI-powered transcription application with speaker diarization and LLM features"
 authors = [
     {name = "OpenTranscribe Contributors"}