fix(examples): guard against empty choices and message=None in LLM calls

[qizwiz]

What

Add explicit guards at three LLM call sites in example scripts before accessing response.choices[0].message.content.

Why

client.chat.completions.create() can return two empty-response shapes:

1. choices = [] — on content-policy rejections, rate-limit errors, or provider failures
2. choices[0].message = None — e.g. Gemini 2.5 Flash (via OpenAI-compatible endpoint) returns HTTP 200 with finish_reason: PROHIBITED_CONTENT and message=None

Both crash with IndexError or AttributeError.

Files changed

File	Fix
docs/getting_started/nebius/reason2/src/cosmos_reason2_tests.py	Guard before response.choices[0].message.content
scripts/examples/reason1/av_video_caption_vqa/model_openai.py	Guard before response.choices[0].message.content
scripts/examples/reason1/temporal_localization/process_video_openai_api.py	Guard before resp.choices[0].message.content

# Before
result = response.choices[0].message.content

# After
if not response.choices or response.choices[0].message is None:
    raise ValueError("LLM returned empty or filtered response")
result = response.choices[0].message.content

Corpus context

Detected by pact (llm_response_unguarded mode), a Z3-verified static analyzer for LLM crash vectors. This pattern was found across 13.8k violations in 800+ repos.

[Copilot]

Pull request overview

This PR hardens three example scripts against OpenAI-compatible chat completion responses that return either an empty choices list or a None message, preventing IndexError/AttributeError crashes when accessing choices[0].message.content.

Changes:

• Add a guard for not response.choices before indexing choices[0].
• Add a guard for choices[0].message is None before reading message.content.
• Raise a clear ValueError when the response is empty/filtered.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.

File	Description
docs/getting_started/nebius/reason2/src/cosmos_reason2_tests.py	Adds response-shape validation before printing/returning the LLM result in the Reason2 test runner.
scripts/examples/reason1/av_video_caption_vqa/model_openai.py	Adds response-shape validation before returning generated text from the OpenAI wrapper.
scripts/examples/reason1/temporal_localization/process_video_openai_api.py	Adds response-shape validation before returning caption/localization output from video processing.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.