feat: add stock_analysis & fix requirements in content_safety (#98)

cuericlee · web-flow · commit 7a695fcb76d4 · 2025-12-26T10:22:39.000+08:00
diff --git a/02-use-cases/beginner/callback/agent.py b/02-use-cases/beginner/callback/agent.py
@@ -1,5 +1,3 @@
-import asyncio
-
 from agentkit.apps import AgentkitAgentServerApp
 from veadk import Agent, Runner
 from veadk.memory import ShortTermMemory
@@ -50,5 +48,5 @@ async def main():
 
 
 if __name__ == "__main__":
-    asyncio.run(main())
-    # agent_server_app.run(host="0.0.0.0", port=8000)
+    # asyncio.run(main())
+    agent_server_app.run(host="0.0.0.0", port=8000)
diff --git a/02-use-cases/content_safety/pyproject.toml b/02-use-cases/content_safety/pyproject.toml
@@ -5,7 +5,7 @@ description = "Add your description here"
 readme = "README.md"
 requires-python = ">=3.12"
 dependencies = [
-    "agentkit-sdk-python==0.2.2",
+    "agentkit-sdk-python>=0.2.0",
     "google-adk==1.19.0",
     "veadk-python==0.5.0",
 ]
diff --git a/02-use-cases/content_safety/requirements.txt b/02-use-cases/content_safety/requirements.txt
@@ -1,3 +1,3 @@
 veadk-python==0.5.0
 google-adk==1.19.0
-agentkit-sdk-python==0.2.2
+agentkit-sdk-python>=0.2.0
diff --git a/02-use-cases/stock_analysis/README.md b/02-use-cases/stock_analysis/README.md
@@ -0,0 +1,224 @@
+# 股票分析智能助手 - Stock Analysis
+
+## 概述
+
+这是一个基于火山引擎AgentKit构建的具备股票分析功能的智能助手。
+
+## 核心功能
+
+- 功能 1：股票数据分析
+- 功能 2：股票走势预测
+- 功能 3：股票投资建议
+
+## Agent 能力
+
+```text
+用户消息
+    ↓
+AgentKit 运行时
+    ↓
+Stock Analysis Agent
+    ├── 网页信息工具 (web_search)
+    ├── 代码执行工具 (run_code)
+```
+
+## 目录结构说明
+
+```bash
+stock_analysis/
+├── agent.py           # Agent
+├── client.py          # 测试客户端（SSE 流式调用）
+├── requirements.txt   # Python 依赖列表 （agentkit部署时需要指定依赖文件)
+├── pyproject.toml     # 项目配置（uv 依赖管理）
+├── agentkit.yaml      # AgentKit 部署配置（运行agentkit config之后会自动生成）
+├── Dockerfile         # Docker 镜像构建文件（运行agentkit config之后会自动生成）
+└── README.md          # 项目说明文档
+```
+
+## 本地运行
+
+### 前置准备
+
+**1. 开通火山方舟模型服务：**
+
+- 访问 [火山方舟控制台](https://exp.volcengine.com/ark?mode=chat)
+- 开通模型服务
+
+**2. 获取火山引擎访问凭证：**
+
+- 参考 [用户指南](https://www.volcengine.com/docs/6291/65568?lang=zh) 获取 AK/SK
+
+**3. 开通`web_search`工具权限：**
+
+- 使用[`web_search`工具](https://www.volcengine.com/docs/85508/1650263)，需提前开通并创建联网问答Agent[相应权限](https://www.volcengine.com/docs/85508/1544858)
+
+**4. 创建 AgentKit 工具：**
+
+- 工具类型选择：预置工具 -> AIO Sandbox
+
+### 依赖安装
+
+#### 1. 安装 uv 包管理器
+
+```bash
+# macOS / Linux（官方安装脚本）
+curl -LsSf https://astral.sh/uv/install.sh | sh
+
+# 或使用 Homebrew（macOS）
+brew install uv
+```
+
+#### 2. 初始化项目依赖
+
+```bash
+# 进入项目目录
+cd 02-use-cases/stock_analysis
+```
+
+您可以通过 `pip` 工具来安装本项目依赖：
+
+```bash
+pip install -r requirements.txt
+```
+
+或者使用 `uv` 工具来安装本项目依赖：
+
+```bash
+# 如果没有 `uv` 虚拟环境，可以使用命令先创建一个虚拟环境
+uv venv --python 3.12
+
+# 使用 `pyproject.toml` 管理依赖
+uv sync --index-url https://pypi.tuna.tsinghua.edu.cn/simple
+
+# 使用 `requirements.txt` 管理依赖
+uv pip install -r requirements.txt
+
+# 激活虚拟环境
+source .venv/bin/activate
+```
+
+### 环境准备
+
+```bash
+# 火山方舟模型名称
+export MODEL_AGENT_NAME=doubao-seed-1-6-251015
+
+# 火山引擎访问凭证（必需）
+export VOLCENGINE_ACCESS_KEY=<Your Access Key>
+export VOLCENGINE_SECRET_KEY=<Your Secret Key>
+
+export AGENTKIT_TOOL_ID=<Your Tool ID>
+```
+
+### 调试方法
+
+#### 单线程运行：使用 VeADK Web 调试界面，调试 agent.py
+
+```bash
+# 进入 02-use-cases 目录
+cd agentkit-samples/02-use-cases
+
+# 启动 VeADK Web 界面
+veadk web --port 8080
+
+# 在浏览器访问：http://127.0.0.1:8080
+```
+
+Web 界面提供图形化对话测试环境，支持实时查看消息流和调试信息。
+
+此外，还可以使用命令行测试，调试 agent.py。
+
+```bash
+cd agentkit-samples/02-use-cases/stock_analysis
+
+# 启动 Agent 服务
+uv run agent.py
+# 服务将监听 http://0.0.0.0:8000
+
+# 新开终端，运行测试客户端
+# 需要编辑 client.py，将其中的第 14 行和第 15 行的 base_url 和 api_key 修改为 agentkit.yaml 中生成的 runtime_endpoint 和 runtime_apikey 字段
+uv run client.py
+```
+
+## AgentKit 部署
+
+### 前置准备
+
+**重要提示**：在运行本示例之前，请先访问 [AgentKit 控制台授权页面](https://console.volcengine.com/agentkit/region:agentkit+cn-beijing/auth?projectName=default) 对所有依赖服务进行授权，确保案例能够正常执行。
+
+**1. 开通火山方舟模型服务：**
+
+- 访问 [火山方舟控制台](https://exp.volcengine.com/ark?mode=chat)
+- 开通模型服务
+
+**2. 获取火山引擎访问凭证：**
+
+- 参考 [用户指南](https://www.volcengine.com/docs/6291/65568?lang=zh) 获取 AK/SK
+
+**3. 开通`web_search`工具权限：**
+
+- 使用[`web_search`工具](https://www.volcengine.com/docs/85508/1650263)，需提前开通并创建联网问答Agent[相应权限](https://www.volcengine.com/docs/85508/1544858)
+
+**4. 创建 AgentKit 工具：**
+
+- 工具类型选择：预置工具 -> AIO Sandbox
+
+**5. 设置环境变量：**
+
+```bash
+# 火山引擎访问凭证（必需）
+export VOLCENGINE_ACCESS_KEY=<Your Access Key>
+export VOLCENGINE_SECRET_KEY=<Your Secret Key>
+```
+
+### AgentKit 云上部署
+
+```bash
+cd agentkit-samples/02-use-cases/stock_analysis
+
+# 配置部署参数
+# optional：如果 agentkit config 中不添加 --runtime_envs AGENTKIT_TOOL_ID={{your_tool_id}}，可以在 AgentKit 控制台 智能体运行时 中，关键组件，选择 沙箱工具，并发布
+agentkit config \
+--agent_name stock_analysis \
+--entry_point 'agent.py' \
+--runtime_envs AGENTKIT_TOOL_ID={{your_tool_id}} \
+--launch_type cloud
+
+# 启动云端服务
+agentkit launch
+
+# 测试部署的 Agent
+agentkit invoke '帮我搜一下宁德时代近期的股票走势, 并给我一个简单的投资建议。'
+
+# 或使用 client.py 连接云端服务
+# 需要编辑 client.py，将其中的第 14 行和第 15 行的 base_url 和 api_key 修改为 agentkit.yaml 中生成的 runtime_endpoint 和 runtime_apikey 字段
+uv run client.py
+```
+
+## 示例提示词
+
+以下是一些常用的提示词示例：
+
+- "帮我搜一下比亚迪近期股票数据，分析一下走势"
+- "帮我搜一下宁德时代近期的股票走势, 并给我一个简单的投资建议"
+
+## 效果展示
+
+| 示例提示词 1 | 示例提示词 2 |
+| -------- | ------- |
+| 帮我搜一下比亚迪近期股票数据，分析一下走势。 | 帮我搜一下宁德时代近期的股票走势, 并给我一个简单的投资建议 |
+| ![示例提示词 1 截图](assets/images/prompt1.jpeg) | ![示例提示词 2 截图](assets/images/prompt2.jpeg) |
+
+## 常见问题
+
+无。
+
+## 参考资料
+
+- [VeADK 官方文档](https://volcengine.github.io/veadk-python/)
+- [AgentKit 开发指南](https://volcengine.github.io/agentkit-sdk-python/)
+- [火山方舟模型服务](https://console.volcengine.com/ark/region:ark+cn-beijing/overview?briefPage=0&briefType=introduce&type=new&projectName=default)
+
+## 代码许可
+
+本工程遵循 Apache 2.0 License
diff --git a/02-use-cases/stock_analysis/agent.py b/02-use-cases/stock_analysis/agent.py
@@ -0,0 +1,47 @@
+# Copyright (c) 2025 Beijing Volcano Engine Technology Co., Ltd. and/or its affiliates.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from veadk import Agent
+from veadk.memory.short_term_memory import ShortTermMemory
+from veadk.tools.builtin_tools.run_code import run_code
+from veadk.tools.builtin_tools.web_search import web_search
+from agentkit.apps import AgentkitAgentServerApp
+
+short_term_memory = ShortTermMemory(backend="local")
+
+agent = Agent(
+    name="data_analysis_agent",
+    description="A data analysis for stock marketing",
+    instruction="""
+    You are a data analysis agent for stock marketing.
+    Talk with user friendly. You can invoke your tools to finish user's task or question.
+    Load memory first. In case you already have answer from memory, you can use it directly.
+    Download the stock data thru sandbox if it is not available in memory.
+    * If trading data is not found, download the stock trading data using run_code. You can use the Python library akshare to download relevant stock data.
+    * After downloading, execute code through run_code to avoid installation checks each time.
+    * You can use the web_search tool to search for relevant company operational data.
+    * If dependency libraries are missing, install them for the sandbox using Python code.
+    """,
+    tools=[run_code, web_search],
+)
+
+root_agent = agent
+
+agent_server_app = AgentkitAgentServerApp(
+    agent=agent,
+    short_term_memory=short_term_memory,
+)
+
+if __name__ == "__main__":
+    agent_server_app.run(host="0.0.0.0", port=8000)
diff --git a/02-use-cases/stock_analysis/assets/images/prompt1.jpeg b/02-use-cases/stock_analysis/assets/images/prompt1.jpeg
diff --git a/02-use-cases/stock_analysis/assets/images/prompt2.jpeg b/02-use-cases/stock_analysis/assets/images/prompt2.jpeg
diff --git a/02-use-cases/stock_analysis/client.py b/02-use-cases/stock_analysis/client.py
@@ -0,0 +1,78 @@
+import requests
+import httpx
+import random
+from google.adk.cli.adk_web_server import CreateSessionRequest, RunAgentRequest
+from google.genai.types import Content, Part
+import asyncio
+
+
+if __name__ == "__main__":
+    # Step 0: setup running configs
+    app_name = "content_safety_app"
+    user_id = "agentkit_user"
+    session_id = "agentkit_sample_session"
+    base_url = "http://127.0.0.1:8000"
+    api_key = "test key"
+
+    task_num = 1
+
+    # Step 1: create a session
+    def create_session():
+        create_session_request = CreateSessionRequest(
+            session_id=session_id + f"_{random.randint(1, 9999)}",
+        )
+
+        response = requests.post(
+            url=f"{base_url}/apps/{app_name}/users/{user_id}/sessions/{create_session_request.session_id}",
+            headers={"Authorization": f"Bearer {api_key}"},
+        )
+
+        print(f"[create session] Response from server: {response.json()}")
+
+        return create_session_request.session_id
+
+    # Step 2: run agent with SSE
+    run_agent_request = RunAgentRequest(
+        app_name=app_name,
+        user_id=user_id,
+        session_id=create_session(),
+        new_message=Content(
+            parts=[
+                Part(text="帮我搜一下宁德时代近期的股票走势, 并给我一个简单的投资建议")
+            ],
+            role="user",
+        ),
+        stream=True,
+    )
+
+    print("[run agent] Event from server:")
+
+    # 3. Handle streaming events
+    async def send_request(message: str):
+        run_agent_request = RunAgentRequest(
+            app_name=app_name,
+            user_id=user_id,
+            session_id=create_session(),
+            new_message=Content(parts=[Part(text=message)], role="user"),
+            stream=True,
+        )
+
+        with httpx.stream(
+            "POST",
+            f"{base_url}/run_sse",
+            json=run_agent_request.model_dump(exclude_none=True),
+            timeout=120,
+            headers={"Authorization": f"Bearer {api_key}"},
+        ) as r:
+            for line in r.iter_lines():
+                print(line)
+
+    async def send_request_parallel():
+        await send_request("帮我搜一下宁德时代近期的股票走势, 并给我一个简单的投资建议")
+        # tasks = [
+        #     send_request("")
+        #     for _ in range(task_num)
+        # ]
+        # await asyncio.gather(*tasks)
+
+    asyncio.run(send_request_parallel())
diff --git a/02-use-cases/stock_analysis/config.yaml.example b/02-use-cases/stock_analysis/config.yaml.example
diff --git a/02-use-cases/stock_analysis/pyproject.toml b/02-use-cases/stock_analysis/pyproject.toml
@@ -0,0 +1,11 @@
+[project]
+name = "stock-analysis"
+version = "0.1.0"
+description = "Add your description here"
+readme = "README.md"
+requires-python = ">=3.12"
+dependencies = [
+    "agentkit-sdk-python>=0.2.0",
+    "google-adk==1.19.0",
+    "veadk-python==0.5.0",
+]
diff --git a/02-use-cases/stock_analysis/requirements.txt b/02-use-cases/stock_analysis/requirements.txt
@@ -0,0 +1,3 @@
+veadk-python==0.5.0
+google-adk==1.19.0
+agentkit-sdk-python>=0.2.0

Original file line number	Diff line number	Diff line change
`@@ -5,7 +5,7 @@ description = "Add your description here"`
`5`	`5`	`readme = "README.md"`
`6`	`6`	`requires-python = ">=3.12"`
`7`	`7`	`dependencies = [`
`8`		`- "agentkit-sdk-python==0.2.2",`
	`8`	`+ "agentkit-sdk-python>=0.2.0",`
`9`	`9`	`"google-adk==1.19.0",`
`10`	`10`	`"veadk-python==0.5.0",`
`11`	`11`	`]`