Adding Support for GPT5.2

[Jack-Khuu]

As titled

Note this doesn't enable reasoningEffort.xhigh as this is currently timingout/pushing token limits

---

list-models

🔹 OPENAI Provider:
--------------------------------------------------
  o4-mini                             ✅ Available
    └─ OpenAI o4-mini - fast reasoning model
  gpt-5                               ✅ Available
    └─ GPT-5 flagship model (Released Aug 2025)
  gpt-5.2                             ✅ Available
    └─ GPT-5.2 flagship model (Released Dec 2025)

Auto Agent
python -m Fuser.auto_agent --problem /home/jackkhuu/KernelBench/KernelBench/level3/1_MLP.py --ka-model gpt-5.2 --router-model gpt-5.2 --extract-model gpt-5.2 --dispatch-model gpt-5.2 --compose-model gpt-5.2 --verify --ignore-router-config