New LM Studio REST API v1

**Describe the bug**
proably the LM Studio provider in Goose AI agent desktop is not updated yet to the newer LM Studio REST API v1, which enables more endpoints than the Open AI compatible one.
This would include stats with total output tokes and capabilities with vision, which are currently not received in Goose AI for qwen 3.8 27b.
In particular the former would trigger the auto-compact feature, which would then not abruptly disconnect the LM Studio server with network error messages.

💡 Before filing, please check common issues:
https://goose-docs.ai/docs/troubleshooting

📦 To help us debug faster, attach your **diagnostics JSON report** if possible.
👉 How to capture it: https://goose-docs.ai/docs/troubleshooting/diagnostics-and-reporting/

A clear and concise description of what the bug is.

---

**To Reproduce**
Steps to reproduce the behavior:
1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

---

**Expected behavior**
A clear and concise description of what you expected to happen.

---

**Screenshots**
If applicable, add screenshots to help explain your problem.

---

**Please provide the following information**
- **OS & Arch:** linux 7.2.3-200.nobara.fc44.x86_64 x86_64
- **Interface:** UI
- **Version:** 1.49.0
- **Extensions enabled:** computercontroller, developer, scheduler, code_execution, Extension Manager, summarize, chatrecall, apps, analyze, skills, summon, autovisualiser, memory, Fetch, Knowledge Graph Memory, Playwright, YouTube Transcript, Tavily Web Search
- **Provider & Model:** lmstudio - qwen/qwen3.8-27b

---

**Additional context**
Add any other context about the problem here.

Look for the LM Studio provider implementation in the codebase, likely under a providers or integrations directory. Compare the current Open AI compatible API calls with the newer LM Studio REST API v1 endpoints. The goal is to update the provider to support new endpoints for stats (total output tokens) and capabilities (vision). Check for existing tests related to LM Studio to ensure compatibility.

Written by the indexing model from the issue text.