fix(stainless): handle [DONE] SSE terminator in streaming responses

[dtmeadows]

Changes

Adds a top-level streaming.on_event config to client-sdks/stainless/config.yml that gracefully handles the OpenAI-standard data: [DONE] SSE stream terminator. Without this, the generated SDKs crash when consuming vLLM or any other OpenAI-compatible backend, because they attempt to JSON-parse the [DONE] sentinel as if it were a regular event chunk — producing errors like Could not parse message into JSON: [DONE].

The fix adds a priority-ordered event handler table. The first rule matches data: [DONE] and signals a clean stream end (handle: done); the fallthrough rule yields all other events normally.

streaming:
  on_event:
    - data_starts_with: "[DONE]"
      handle: done
    - kind: fallthrough
      handle: yield
      error_property: error

Notes

Fixes #4744.

[github-actions]

✱ Stainless preview builds

This PR will update the llama-stack-client SDKs with the following commit message.

fix(stainless): handle [DONE] SSE terminator in streaming responses

Edit this comment to update it. It will appear in the SDK's changelogs.

✅ llama-stack-client-node studio · code · diff

Your SDK built successfully.
generate ⚠️ → build ✅ → lint ✅ → test ❗

npm install https://pkg.stainless.com/s/llama-stack-client-node/3eca3a1cc6115e23766991ad8775670a0386e942/dist.tar.gz

⏳ llama-stack-client-go studio · code · diff

generate ❗ → build ⏳ → lint ❗ → test ❗

go get github.com/stainless-sdks/llama-stack-client-go@67a4b2d546608ce3408df614e6113cdb131a448b

✅ llama-stack-client-python studio · code · diff

Your SDK built successfully.
generate ⚠️ → build ✅ → lint ✅ → test ✅

pip install https://pkg.stainless.com/s/llama-stack-client-python/5ec4518716be86ac2444d480665068fb3b4b1118/llama_stack_client-0.5.0a2-py3-none-any.whl

New diagnostics (2 note)

💡 ReadmeExample/MissingParam: Example is missing required parameter `messages`.

💡 ReadmeExample/MissingParam: Example is missing required parameter `model`.

✅ llama-stack-client-openapi studio · code · diff

Your SDK built successfully.
generate ⚠️

⏳ These are partial results; builds are still running.

---

This comment is auto-generated by GitHub Actions and is automatically kept up to date as you push.
If you push custom code to the preview branch, re-run this workflow to update the comment.
Last updated: 2026-02-27 19:25:11 UTC

[mattf]

before -

$ uv run --with llama-stack-client python -c 'from llama_stack_client import LlamaStackClient; client = LlamaStackClient(base_url="http://localhost:8000"); print([chunk for chunk in client.chat.completions.create(model="Qwen/Qwen3-0.6B", messages=[{"role": "user", "content": "hello"}], max_tokens=5, stream=True)])'
INFO:httpx:HTTP Request: POST http://localhost:8000/v1/chat/completions "HTTP/1.1 200 OK"
...
json.decoder.JSONDecodeError: Expecting value: line 1 column 2 (char 1)

after -

$ uv run --with https://pkg.stainless.com/s/llama-stack-client-python/e8ba87ec863e1b417327c14b112fbd8622505dbc/llama_stack_client-0.5.0a2-py3-none-any.whl python -c 'from llama_stack_client import LlamaStackClient; client = LlamaStackClient(base_url="http://localhost:8000"); print([chunk for chunk in client.chat.completions.create(model="Qwen/Qwen3-0.6B", messages=[{"role": "user", "content": "hello"}], max_tokens=5, stream=True)])'       
...
[ChatCompletionChunk(id='chatcmpl-a9c1fa7389f8f8f9', choices=[Choice(delta=ChoiceDelta(content='', reasoning_content=None, refusal=None, role='assistant', tool_calls=None), index=0, finish_reason=None, logprobs=None)], created=1772205573, model='Qwen/Qwen3-0.6B', object='chat.completion.chunk', service_tier=None, usage=None, prompt_token_ids=None), ChatCompletionChunk(id='chatcmpl-a9c1fa7389f8f8f9', choices=[Choice(delta=ChoiceDelta(content=None, reasoning_content=None, refusal=None, role=None, tool_calls=None, reasoning='\n'), index=0, finish_reason=None, logprobs=None, token_ids=None)], created=1772205573, model='Qwen/Qwen3-0.6B', object='chat.completion.chunk', service_tier=None, usage=None), ChatCompletionChunk(id='chatcmpl-a9c1fa7389f8f8f9', choices=[Choice(delta=ChoiceDelta(content=None, reasoning_content=None, refusal=None, role=None, tool_calls=None, reasoning='Okay'), index=0, finish_reason=None, logprobs=None, token_ids=None)], created=1772205573, model='Qwen/Qwen3-0.6B', object='chat.completion.chunk', service_tier=None, usage=None), ChatCompletionChunk(id='chatcmpl-a9c1fa7389f8f8f9', choices=[Choice(delta=ChoiceDelta(content=None, reasoning_content=None, refusal=None, role=None, tool_calls=None, reasoning=','), index=0, finish_reason=None, logprobs=None, token_ids=None)], created=1772205573, model='Qwen/Qwen3-0.6B', object='chat.completion.chunk', service_tier=None, usage=None), ChatCompletionChunk(id='chatcmpl-a9c1fa7389f8f8f9', choices=[Choice(delta=ChoiceDelta(content=None, reasoning_content=None, refusal=None, role=None, tool_calls=None, reasoning=' the'), index=0, finish_reason='length', logprobs=None, stop_reason=None, token_ids=None)], created=1772205573, model='Qwen/Qwen3-0.6B', object='chat.completion.chunk', service_tier=None, usage=None)]