feat: 添加 Google ADK、CrewAI、LangChain 和 PydanticAI 模型适配器的流选项支持，更新数据 AP…

.github/workflows/release-test.yml

@@ -0,0 +1,135 @@
+name: Release Test Package
+
+on:
+  # 支持手动触发，用户可以在 GitHub Actions 页面点击 "Run workflow"
+  workflow_dispatch:
+    inputs:
+      version_bump:
+        description: '版本递增类型'
+        required: true
+        default: 'patch'
+        type: choice
+        options:
+          - patch   # 0.0.1 -> 0.0.2
+          - minor   # 0.0.1 -> 0.1.0
+          - major   # 0.0.1 -> 1.0.0
+
+jobs:
+  release-test:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+      id-token: write
+    steps:
+    - uses: actions/checkout@v4
+      with:
+        fetch-depth: 0  # 获取所有历史和 tags
+
+    - name: Set up Python
+      uses: actions/setup-python@v5
+      with:
+        python-version: '3.10'
+
+    - name: Get latest test version and calculate next version
+      id: version
+      run: |
+        # 获取所有 agentrun-inner-test-v* 的 tags，找到最新版本
+        LATEST_TAG=$(git tag -l "agentrun-inner-test-v*" | sort -V | tail -n 1)
+
+        if [ -z "$LATEST_TAG" ]; then
+          # 如果没有找到任何 tag，从 0.0.0 开始
+          CURRENT_VERSION="0.0.0"
+          echo "No existing test tags found, starting from 0.0.0"
+        else
+          # 从 tag 中提取版本号
+          CURRENT_VERSION="${LATEST_TAG#agentrun-inner-test-v}"
+          echo "Latest test tag: $LATEST_TAG (version: $CURRENT_VERSION)"
+        fi
+
+        # 解析版本号
+        IFS='.' read -r MAJOR MINOR PATCH <<< "$CURRENT_VERSION"
+
+        # 根据用户选择递增版本
+        BUMP_TYPE="${{ inputs.version_bump }}"
+        case "$BUMP_TYPE" in
+          major)
+            MAJOR=$((MAJOR + 1))
+            MINOR=0
+            PATCH=0
+            ;;
+          minor)
+            MINOR=$((MINOR + 1))
+            PATCH=0
+            ;;
+          patch)
+            PATCH=$((PATCH + 1))
+            ;;
+        esac
+
+        NEW_VERSION="${MAJOR}.${MINOR}.${PATCH}"
+        NEW_TAG="agentrun-inner-test-v${NEW_VERSION}"
+
+        echo "VERSION=${NEW_VERSION}" >> $GITHUB_OUTPUT
+        echo "TAG=${NEW_TAG}" >> $GITHUB_OUTPUT
+        echo "New version: ${NEW_VERSION}"
+        echo "New tag: ${NEW_TAG}"
+
+    - name: Update package name and version in pyproject.toml
+      run: |
+        VERSION="${{ steps.version.outputs.VERSION }}"
+        # 修改包名为 agentrun-inner-test
+        sed -i 's/name = "agentrun-sdk"/name = "agentrun-inner-test"/' pyproject.toml
+        # 修改版本号
+        sed -i 's/version = "[^"]*"/version = "'${VERSION}'"/' pyproject.toml
+        echo "Updated pyproject.toml:"
+        head -10 pyproject.toml
+
+    - name: Update __version__ in __init__.py
+      run: |
+        VERSION="${{ steps.version.outputs.VERSION }}"
+        if grep -q "__version__" agentrun/__init__.py; then
+          sed -i 's/__version__ = "[^"]*"/__version__ = "'${VERSION}'"/' agentrun/__init__.py
+        else
+          sed -i '1a __version__ = "'${VERSION}'"' agentrun/__init__.py
+        fi
+        echo "Updated __init__.py version to ${VERSION}"
+        grep "__version__" agentrun/__init__.py
+
+    - name: Build package
+      run: |
+        python -m pip install --upgrade pip
+        pip install build twine
+        python -m build
+        echo "Package built successfully"
+        ls -la dist/
+
+    - name: Verify package
+      run: |
+        python -m twine check dist/*
+        echo "Package verification completed"
+
+    - name: Publish to PyPI
+      uses: pypa/gh-action-pypi-publish@release/v1
+      with:
+        password: ${{ secrets.PYPI_API_TOKEN }}
+        verify-metadata: false
+
+    - name: Create and push tag
+      run: |
+        TAG="${{ steps.version.outputs.TAG }}"
+        git config --local user.email "action@github.com"
+        git config --local user.name "GitHub Action"
+        git tag -a "$TAG" -m "Release test package version ${{ steps.version.outputs.VERSION }}"
+        git push origin "$TAG"
+        echo "Created and pushed tag: $TAG"
+
+    - name: Summary
+      run: |
+        echo "## 🎉 Test Package Released!" >> $GITHUB_STEP_SUMMARY
+        echo "" >> $GITHUB_STEP_SUMMARY
+        echo "- **Package Name:** agentrun-inner-test" >> $GITHUB_STEP_SUMMARY
+        echo "- **Version:** ${{ steps.version.outputs.VERSION }}" >> $GITHUB_STEP_SUMMARY
+        echo "- **Tag:** ${{ steps.version.outputs.TAG }}" >> $GITHUB_STEP_SUMMARY
+        echo "- **Commit:** ${{ github.sha }}" >> $GITHUB_STEP_SUMMARY
+        echo "" >> $GITHUB_STEP_SUMMARY
+        echo "Install with: \`pip install agentrun-inner-test==${{ steps.version.outputs.VERSION }}\`" >> $GITHUB_STEP_SUMMARY

agentrun/integration/agentscope/model_adapter.py

@@ -56,4 +56,5 @@ def wrap_model(self, common_model: CommonModel) -> Any:
                 "base_url": info.base_url,
                 "http_client": AsyncClient(headers=info.headers),
             },
+            generate_kwargs={"stream_options": {"include_usage": True}},
         )

agentrun/integration/crewai/model_adapter.py

@@ -22,5 +22,6 @@ def wrap_model(self, common_model: Any) -> Any:
             model=f"{info.provider or 'openai'}/{info.model}",
             base_url=info.base_url,
             default_headers=info.headers,
+            stream_options={"include_usage": True},
             # async_client=AsyncClient(headers=info.headers),
         )

agentrun/integration/google_adk/model_adapter.py

@@ -39,4 +39,5 @@ def wrap_model(self, common_model: CommonModel) -> Any:
             api_base=info.base_url,
             api_key=info.api_key,
             extra_headers=info.headers,
+            stream_options={"include_usage": True},
         )

agentrun/integration/langchain/model_adapter.py

@@ -33,4 +33,5 @@ def wrap_model(self, common_model: Any) -> Any:
             model=info.model,
             base_url=info.base_url,
             async_client=AsyncClient(headers=info.headers),
+            stream_usage=True,
         )

agentrun/integration/pydantic_ai/model_adapter.py

@@ -1,8 +1,6 @@
 """PydanticAI 模型适配器 / PydanticAI Model Adapter"""
 
-from contextlib import asynccontextmanager
-import json
-from typing import Any, AsyncIterator
+from typing import Any
 
 from agentrun.integration.utils.adapter import ModelAdapter
 from agentrun.integration.utils.model import CommonModel
@@ -19,6 +17,7 @@ def wrap_model(self, common_model: CommonModel) -> Any:
         try:
             from pydantic_ai.models.openai import OpenAIChatModel
             from pydantic_ai.providers.openai import OpenAIProvider
+            from pydantic_ai.settings import ModelSettings
         except Exception as e:
             raise ImportError(
                 "PydanticAI is not installed. "
@@ -36,6 +35,9 @@ def wrap_model(self, common_model: CommonModel) -> Any:
                 api_key=info.api_key,
                 http_client=AsyncClient(headers=info.headers),
             ),
+            settings=ModelSettings(
+                extra_body={"stream_options": {"include_usage": True}}
+            ),
         )
 
 

agentrun/model/api/data.py

@@ -54,6 +54,10 @@ def completions(
             **self.headers,
             **kwargs.get("headers", {}),
         }
+        if kwargs.get("stream_options") is None:
+            kwargs["stream_options"] = {}
+        kwargs["stream_options"]["include_usage"] = True
+
         from litellm import completion
 
         return completion(
@@ -82,6 +86,9 @@ def responses(
             **self.headers,
             **kwargs.get("headers", {}),
         }
+        if kwargs.get("stream_options") is None:
+            kwargs["stream_options"] = {}
+        kwargs["stream_options"]["include_usage"] = True
         from litellm import responses
 
         return responses(

tests/unittests/integration/test_integration.py

@@ -53,12 +53,16 @@ def fake_completion(*args, **kwargs):
             messages = kwargs.get("messages") or []
             tools_payload = kwargs.get("tools")
             assert kwargs.get("stream") in (None, False)
+            assert pydash.get(kwargs, "stream_options.include_usage") is True
+
             return self._build_model_response(messages, tools_payload)
 
         async def fake_acompletion(*args, **kwargs):
             messages = kwargs.get("messages") or []
             tools_payload = kwargs.get("tools")
             assert kwargs.get("stream") in (None, False)
+            assert pydash.get(kwargs, "stream_options.include_usage") is True
+
             return self._build_model_response(messages, tools_payload)
 
         monkeypatch.setattr("litellm.completion", fake_completion)