Add GPT-5.1 family support

[PeterDaveHello]

Summary from GitHub Copilot:

This pull request adds support for the new GPT-5.1 family of models throughout the codebase, including constants, configuration, and UI options. It also ensures that model alias resolution and context lengths are handled for these new models, and adds a test to verify alias resolution.

Support for GPT-5.1 models:

• Added new constants for GPT_5_1_CHAT_LATEST_MODEL, GPT_5_1_MODEL, GPT_5_1_2025_11_13_MODEL, GPT_5_1_CODEX_MODEL, and GPT_5_1_CODEX_MINI_MODEL in app/openai_constants.py.
• Registered these new models in the versioning, context length, and alias resolution mappings in app/openai_constants.py. [1] [2] [3]

Slack UI integration:

• Imported the new GPT-5.1 model constants in app/slack_ui.py and added them as selectable options in the model configuration modal. [1] [2]

Testing:

• Added a test in tests/model_constants_test.py to ensure that the GPT_5_1_MODEL alias resolves correctly to the dated release.

[chatgpt-codex-connector]

💡 Codex Review

ChatGPT-in-Slack/app/openai_ops.py

Lines 86 to 102 in a632335

	def _is_reasoning(model: str) -> bool:
	"""Returns True if the model is a reasoning model under Chat Completions.
	Excludes chat models like gpt-5-chat-latest and gpt-5-search-api. Matches o3*, o4*, and
	non-chat gpt-5* families. Case-insensitive and safe with None/empty.
	"""
	if not model:
	return False
	ml = model.lower()
	if ml.startswith("gpt-5-chat") or ml.startswith("gpt-5-search"):
	return False
	return (
	ml.startswith("o1")
	or ml.startswith("o3")
	or ml.startswith("o4")
	or ml.startswith("gpt-5")
	)

Update reasoning check for gpt-5.1 chat models

The _is_reasoning helper still only excludes names starting with gpt-5-chat or gpt-5-search. The newly added gpt-5.1-chat-latest option does not match those prefixes, so it is now classified as a reasoning model. When users pick this chat model, requests will be sent with max_completion_tokens and without temperature, which is the reasoning-specific payload and can cause parameter errors or ignore user-provided temperature. The heuristic should explicitly exempt gpt-5.1-chat-* (and any future numbered chat variants) the same way it already exempts gpt-5-chat-*.

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".