OpenAI collect chat completion at done

[lrafeei]

OpenAI reintroduced logic that closes the streaming response even if the consumer hasn't read all data. The Python Agent previously relied on reaching the stream's end before sending LLM data to the collector. This resulted in the chat completion data not being collected at the closing of the stream. Logic has now been added to account for this.

(This branch also includes the Starlette StreamingResponse context propagation fix.)

[github-actions]

✅MegaLinter analysis: Success

Descriptor	Linter	Files	Fixed	Errors	Warnings	Elapsed time
✅ ACTION	actionlint	8		0	0	0.8s
✅ MARKDOWN	markdownlint	7	0	0	0	1.11s
✅ PYTHON	ruff	1030	2	0	0	1.05s
✅ PYTHON	ruff-format	1030	3	0	0	0.36s
✅ YAML	prettier	19	0	0	0	1.3s
✅ YAML	v8r	19		0	0	5.7s
✅ YAML	yamllint	19		0	0	0.65s

See detailed reports in MegaLinter artifacts

Show us your support by starring ⭐ the repository