{"repo": "bcefghj/multi-agent-meeting-assistant", "files": {".env.example": {"dbb721ce896829cd38ed2310326231323d732bef": {"t": 1785604099.0141478, "n": 1120, "h": "c0f2d6c95b69d0c0", "c": "# ============================================\n# 多Agent智能会议助手系统 - 环境变量配置模板\n# 复制此文件为 .env 并填入真实值\n# cp .env.example .env\n# ============================================\n\n# ---------- LLM API ----------\n# MiniMax API (推荐，支持中文)\nMINIMAX_API_KEY=your_minimax_api_key_here\nMINIMAX_GROUP_ID=your_group_id_here\nMINIMAX_MODEL=abab6.5s-chat\n\n# OpenAI API (备选)\nOPENAI_API_KEY=your_openai_api_key_here\nOPENAI_MODEL=gpt-4o\n\n# ---------- 语音转写 ----------\n# Whisper 配置\nWHISPER_MODEL_SIZE=large-v2\nWHISPER_LANGUAGE=zh\nWHISPER_DEVICE=cuda\n\n# ---------- Jira 集成 ----------\nJIRA_SERVER=https://your-company.atlassian.net\nJIRA_EMAIL=your_email@company.com\nJIRA_API_TOKEN=your_jira_api_token\nJIRA_PROJECT_KEY=MEET\n\n# ---------- 飞书集成 ----------\nFEISHU_APP_ID=your_feishu_app_id\nFEISHU_APP_SECRET=your_feishu_app_secret\nFEISHU_WEBHOOK_URL=https://open.feishu.cn/open-apis/bot/v2/hook/your_webhook_token\n\n# ---------- 数据库 ----------\nDATABASE_URL=postgresql://postgres:password@localhost:5432/meeting_assistant\nREDIS_URL=redis://localhost:6379/0\n\n# ---------- 服务配置 ----------\nSERVER_HOST=0.0.0.0\nSERVER_PORT=8000\nWEBSOCKET_PORT=8001\nLOG_LEVEL=INFO\n"}}, "python/src/integrations/minimax_client.py": {"53ccea64570980d2b7acca18ada91cf395079b8f": {"t": 1785674030.3280056, "n": 3104, "h": "649e3a9ab060ca35", "c": "\"\"\"MiniMax LLM 客户端 - 支持 M2.7 模型调用\"\"\"\n\nfrom __future__ import annotations\n\nimport json\nimport os\nfrom typing import Any\n\nimport httpx\nfrom loguru import logger\nfrom tenacity import retry, stop_after_attempt, wait_exponential\n\n\nclass MiniMaxClient:\n    \"\"\"\n    MiniMax API 客户端\n\n    支持 MiniMax M2.7 模型，兼容 OpenAI 接口格式。\n    文档: https://platform.minimaxi.com/document/guides/chat-model/chat/api\n    \"\"\"\n\n    BASE_URL = \"https://api.minimax.chat/v1\"\n\n    def __init__(\n        self,\n        api_key: str | None = None,\n        group_id: str | None = None,\n        model: str = \"abab6.5s-chat\",\n    ):\n        self.api_key = api_key or os.getenv(\"MINIMAX_API_KEY\", \"\")\n        self.group_id = group_id or os.getenv(\"MINIMAX_GROUP_ID\", \"\")\n        self.model = model\n        self._client = httpx.AsyncClient(\n            timeout=60.0,\n            headers={\"Authorization\": f\"Bearer {self.api_key}\"},\n        )\n\n    @retry(stop=stop_after_attempt(3), wait=wait_exponential(min=1, max=10))\n    async def chat(\n        self,\n        messages: list[dict[str, str]],\n        temperature: float = 0.7,\n        max_tokens: int = 4096,\n        response_format: dict | None = None,\n    ) -> str:\n        \"\"\"\n        调用 MiniMax 聊天接口\n\n        Args:\n            messages: 消息列表 [{\"role\": \"user\", \"content\": \"...\"}]\n            temperature: 温度参数\n            max_tokens: 最大生成token数\n            response_format: 输出格式约束 (如 {\"type\": \"json_object\"})\n\n        Returns:\n            模型生成的文本\n        \"\"\"\n        payload: dict[str, Any] = {\n            \"model\": self.model,\n            \"messages\": messages,\n            \"temperature\": temperature,\n            \"max_tokens\": max_tokens,\n        }\n        if response_format:\n            payload[\"response_format\"] = response_format\n\n        url = f\"{self.BASE_URL}/text/chatcompletion_v2\"\n        if self.group_id:\n            url = f\"{url}?GroupId={self.group_id}\"\n\n        response = await self._client.post(url, json=payload)\n        response.raise_for_status()\n\n        data = response.json()\n        if \"choices\" in data and len(data[\"choices\"]) > 0:\n            return data[\"choices\"][0][\"message\"][\"content\"]\n\n        logger.error(f\"MiniMax API unexpected response: {data}\")\n        raise ValueError(f\"Unexpected API response: {data}\")\n\n    async def chat_json(\n        self,\n        messages: list[dict[str, str]],\n        temperature: float = 0.3,\n        max_tokens: int = 4096,\n    ) -> dict:\n        \"\"\"调用聊天接口并解析 JSON 输出\"\"\"\n        text = await self.chat(\n            messages=messages,\n            temperature=temperature,\n            max_tokens=max_tokens,\n            response_format={\"type\": \"json_object\"},\n        )\n        try:\n            return json.loads(text)\n        except json.JSONDecodeError:\n            start = text.find(\"{\")\n            end = text.rfind(\"}\") + 1\n            if start >= 0 and end > start:\n                return json.loads(text[start:end])\n            raise\n\n    async def close(self):\n        await self._client.aclose()\n\n    async def __aenter__(self):\n        return self\n\n    async def __aexit__(self, *args):\n        await self.close()\n"}}, "docs/tutorial/02-first-agent.md": {"53ccea64570980d2b7acca18ada91cf395079b8f": {"t": 1785674030.6217542, "n": 1824, "h": "897b3ec8fd02a13a", "c": "# 教程 02 - 第一个 Agent\n\n> 用 30 行代码写一个最简单的 Agent\n\n---\n\n## 目标\n\n写一个能自动总结文本的 Agent。输入一段文字，Agent 调用 LLM 返回摘要。\n\n---\n\n## 代码实现\n\n创建文件 `my_first_agent.py`:\n\n```python\n\"\"\"我的第一个 Agent - 文本摘要\"\"\"\n\nimport os\nimport httpx\nfrom dotenv import load_dotenv\n\nload_dotenv()\n\nclass SummaryAgent:\n    \"\"\"一个最简单的 Agent：接收文本，返回摘要\"\"\"\n\n    def __init__(self):\n        self.api_key = os.getenv(\"MINIMAX_API_KEY\", \"\")\n        self.base_url = \"https://api.minimax.chat/v1\"\n\n    def summarize(self, text: str) -> str:\n        \"\"\"调用 LLM 生成摘要\"\"\"\n        response = httpx.post(\n            f\"{self.base_url}/text/chatcompletion_v2\",\n            headers={\"Authorization\": f\"Bearer {self.api_key}\"},\n            json={\n                \"model\": \"abab6.5s-chat\",\n                \"messages\": [\n                    {\"role\": \"system\", \"content\": \"你是一个摘要助手，用3句话总结用户输入的文本。\"},\n                    {\"role\": \"user\", \"content\": text}\n                ],\n                \"temperature\": 0.3,\n            },\n            timeout=30,\n        )\n        data = response.json()\n        return data[\"choices\"][0][\"message\"][\"content\"]\n\n\nif __name__ == \"__main__\":\n    agent = SummaryAgent()\n\n    text = \"\"\"\n    今天的Q3预算评审会议讨论了三个主要议题。\n    首先，Q2预算执行率为87%，研发投入占比42%。\n    其次，Q3计划将预算上调15%，主要投入AI基础设施和人才招聘。\n    王芳建议招聘3名高级算法工程师，年薪80万。\n    最后，赵伟正在对比服务器供应商，下周一出采购方案。\n    \"\"\"\n\n    print(\"原文:\", text)\n    print(\"\\n摘要:\", agent.summarize(text))\n```\n\n运行：\n```bash\npython my_first_agent.py\n```\n\n---\n\n## 代码解读\n\n1. **`__init__`**: 初始化 Agent，加载 API Key\n2. **`summarize`**: Agent 的核心能力——调用 LLM 生成摘要\n3. **System Prompt**: 定义 Agent 的角色和行为规则\n4. **Temperature 0.3**: 低随机性，摘要需要稳定输出\n\n---\n\n## 进化：让 Agent 更强大\n\n这只是最基础的 Agent。在后续教程中，我们将逐步添加：\n- **结构化输出**: JSON Schema 约束（教程03）\n- **多Agent协作**: LangGraph 编排（教程03）\n- **工具调用**: Jira/飞书集成（教程04）\n- **实时处理**: WebSocket + 流式（教程04）\n\n---\n\n下一篇: [03 - 多Agent编排](03-multi-agent.md)\n"}}}}