From 386c8130ea3aea5a73ad4c2f7233a0d2614b5e16 Mon Sep 17 00:00:00 2001
From: m4 <m4@m4deMac-mini.local>
Date: Thu, 20 Aug 2026 09:11:10 +0800
Subject: [PATCH] feat: add streaming _astream to gateway proxy chat model

---
 EvoScientist/llm/gateway_proxy.py | 49 +++++++++++++++++++++++-
 tests/test_gateway_proxy.py       | 63 +++++++++++++++++++++++++++++++
 2 files changed, 111 insertions(+), 1 deletion(-)
 create mode 100644 tests/test_gateway_proxy.py

diff --git a/EvoScientist/llm/gateway_proxy.py b/EvoScientist/llm/gateway_proxy.py
index c0cf3ae..75531d9 100644
--- a/EvoScientist/llm/gateway_proxy.py
+++ b/EvoScientist/llm/gateway_proxy.py
@@ -2,13 +2,14 @@
 
 from __future__ import annotations
 
+import json
 from collections.abc import Mapping, Sequence
 from typing import Any
 
 import httpx
 from langchain_core.language_models.chat_models import BaseChatModel
 from langchain_core.messages import BaseMessage, messages_from_dict, messages_to_dict
-from langchain_core.outputs import ChatGeneration, ChatResult
+from langchain_core.outputs import ChatGeneration, ChatGenerationChunk, ChatResult
 from langchain_core.tools import BaseTool
 from langchain_core.utils.function_calling import convert_to_openai_tool
 from pydantic import Field
@@ -76,6 +77,52 @@ class GatewayProxyChatModel(BaseChatModel):
             raise RuntimeError("AI4SCI_MODEL_PROXY_RESPONSE_INVALID")
         return ChatResult(generations=[ChatGeneration(message=parsed[0])])
 
+    async def _astream(
+        self,
+        messages: list[BaseMessage],
+        stop: list[str] | None = None,
+        run_manager: Any = None,
+        **kwargs: Any,
+    ):
+        del stop, kwargs
+        attempt_id = str(getattr(run_manager, "run_id", None) or self.run_id)
+        payload = {
+            "run_id": self.run_id,
+            "attempt_id": attempt_id,
+            "envelope_signature": self.envelope_signature,
+            "messages": messages_to_dict(messages),
+            "tools": self.bound_tools,
+            "tool_choice": self.bound_tool_choice,
+            "stream": True,
+        }
+        async with httpx.AsyncClient(timeout=httpx.Timeout(660.0, connect=5.0)) as client:
+            async with client.stream(
+                "POST",
+                f"{self.gateway_url.rstrip('/')}/api/internal/recoverable-runs/model/stream",
+                json=payload,
+            ) as response:
+                response.raise_for_status()
+                async for line in response.aiter_lines():
+                    if not line.startswith("data:"):
+                        continue
+                    data = line[len("data:"):].strip()
+                    if data == "[DONE]":
+                        break
+                    chunk = json.loads(data)
+                    message = _chunk_to_message(chunk)
+                    yield ChatGenerationChunk(
+                        message=message,
+                        generation_info=chunk.get("generation_info"),
+                    )
+
+
+def _chunk_to_message(chunk: dict[str, Any]) -> BaseMessage:
+    delta = chunk.get("delta") or {}
+    parsed = messages_from_dict(
+        [delta.get("message") or {"type": "AIMessageChunk", "data": {"content": delta.get("text", "")}}]
+    )
+    return parsed[0]
+
 
 def proxy_from_config(
     value: Mapping[str, Any], *, provider_id: str = "", model_id: str = ""
diff --git a/tests/test_gateway_proxy.py b/tests/test_gateway_proxy.py
new file mode 100644
index 0000000..6107156
--- /dev/null
+++ b/tests/test_gateway_proxy.py
@@ -0,0 +1,63 @@
+import httpx
+import pytest
+from langchain_core.messages import HumanMessage
+
+from EvoScientist.llm.gateway_proxy import GatewayProxyChatModel
+
+
+class _FakeStream:
+    def __init__(self, lines):
+        self._lines = lines
+
+    async def __aenter__(self):
+        return self
+
+    async def __aexit__(self, *exc):
+        return False
+
+    def raise_for_status(self):
+        pass
+
+    async def aiter_lines(self):
+        for line in self._lines:
+            yield line
+
+
+class _FakeClient:
+    def __init__(self, lines):
+        self._lines = lines
+        self.sent = None
+
+    async def __aenter__(self):
+        return self
+
+    async def __aexit__(self, *exc):
+        return False
+
+    def stream(self, method, url, json=None):
+        assert method == "POST"
+        assert url.endswith("/api/internal/recoverable-runs/model/stream")
+        self.sent = json
+        return _FakeStream(self._lines)
+
+
+@pytest.mark.anyio
+async def test_astream_yields_chunks_from_sse(monkeypatch):
+    model = GatewayProxyChatModel(
+        gateway_url="http://gw",
+        run_id="run-1",
+        envelope_signature="sig",
+    )
+    lines = [
+        'data: {"delta": {"text": "hello"}}\n',
+        'data: {"delta": {"text": " world"}}\n',
+        "data: [DONE]\n",
+    ]
+    fake = _FakeClient(lines)
+    monkeypatch.setattr(httpx, "AsyncClient", lambda **kw: fake)
+
+    chunks = [c async for c in model._astream([HumanMessage(content="hi")])]
+
+    assert fake.sent["stream"] is True
+    assert len(chunks) == 2
+    assert chunks[0].message.content == "hello"