Undefined-bot 2.1.0__py3-none-any.whl

Undefined/skills/agents/file_analysis_agent/tools/analyze_multimodal/handler.py

@@ -0,0 +1,178 @@
+from pathlib import Path
+from typing import Any, Dict
+import base64
+import logging
+import aiofiles
+
+logger = logging.getLogger(__name__)
+
+
+# 媒体类型的 MIME 映射表
+MIME_TYPES_MAP = {
+    "image": {
+        ".jpg": "image/jpeg",
+        ".jpeg": "image/jpeg",
+        ".png": "image/png",
+        ".gif": "image/gif",
+        ".bmp": "image/bmp",
+        ".webp": "image/webp",
+        ".svg": "image/svg+xml",
+        ".ico": "image/x-icon",
+    },
+    "audio": {
+        ".mp3": "audio/mpeg",
+        ".wav": "audio/wav",
+        ".flac": "audio/flac",
+        ".aac": "audio/aac",
+        ".m4a": "audio/mp4",
+        ".ogg": "audio/ogg",
+        ".wma": "audio/x-ms-wma",
+    },
+    "video": {
+        ".mp4": "video/mp4",
+        ".avi": "video/x-msvideo",
+        ".mov": "video/quicktime",
+        ".mkv": "video/x-matroska",
+        ".webm": "video/webm",
+        ".flv": "video/x-flv",
+        ".wmv": "video/x-ms-wmv",
+    },
+}
+
+# 默认 MIME 类型
+DEFAULT_MIMES = {
+    "image": "image/jpeg",
+    "audio": "audio/mpeg",
+    "video": "video/mp4",
+}
+
+# 扩展名分组
+EXTENSION_GROUPS = {
+    "image": [
+        ".jpg",
+        ".jpeg",
+        ".png",
+        ".gif",
+        ".bmp",
+        ".webp",
+        ".svg",
+        ".ico",
+    ],
+    "audio": [".mp3", ".wav", ".flac", ".aac", ".m4a", ".ogg", ".wma"],
+    "video": [".mp4", ".avi", ".mov", ".mkv", ".webm", ".flv", ".wmv"],
+}
+
+
+async def execute(args: Dict[str, Any], context: Dict[str, Any]) -> str:
+    file_path: str = args.get("file_path", "")
+    media_type: str = args.get("media_type", "auto")
+    prompt_extra: str = args.get("prompt", "")
+
+    path = Path(file_path)
+
+    if not path.exists():
+        return f"错误：文件不存在 {file_path}"
+
+    if not path.is_file():
+        return f"错误：{file_path} 不是文件"
+
+    ai_client = context.get("ai_client")
+    if not ai_client:
+        return "错误：AI client 未在上下文中提供"
+
+    try:
+        detected_type = _detect_media_type(path, media_type)
+
+        with open(path, "rb") as f:
+            media_data = base64.b64encode(f.read()).decode()
+
+        mime_type = _get_mime_type(detected_type, path)
+        media_content = f"data:{mime_type};base64,{media_data}"
+
+        async with aiofiles.open(
+            "res/prompts/analyze_multimodal.txt", "r", encoding="utf-8"
+        ) as f:
+            prompt = await f.read()
+
+        if prompt_extra:
+            prompt += f"\n\n【补充指令】\n{prompt_extra}"
+
+        content_items: list[dict[str, Any]] = [{"type": "text", "text": prompt}]
+
+        if detected_type == "image":
+            content_items.append(
+                {"type": "image_url", "image_url": {"url": media_content}}
+            )
+        elif detected_type == "audio":
+            content_items.append(
+                {"type": "audio_url", "audio_url": {"url": media_content}}
+            )
+        elif detected_type == "video":
+            content_items.append(
+                {"type": "video_url", "video_url": {"url": media_content}}
+            )
+
+        response = await ai_client._http_client.post(
+            ai_client.vision_config.api_url,
+            headers={
+                "Authorization": f"Bearer {ai_client.vision_config.api_key}",
+                "Content-Type": "application/json",
+            },
+            json=ai_client._build_request_body(
+                model_config=ai_client.vision_config,
+                messages=[{"role": "user", "content": content_items}],
+                max_tokens=8192,
+            ),
+        )
+        response.raise_for_status()
+        result = response.json()
+
+        content = ai_client._extract_choices_content(result)
+        return str(content) if content else "分析失败"
+
+    except Exception as e:
+        logger.exception(f"多模态分析失败: {e}")
+        return f"多模态分析失败: {e}"
+
+
+def _detect_media_type(path: Path, media_type: str) -> str:
+    """根据文件扩展名检测媒体类型
+
+    参数:
+        path: 文件路径
+        media_type: 用户指定的媒体类型（如果是 "auto" 则自动检测）
+
+    返回:
+        检测到的媒体类型 ("image", "audio", "video" 或默认 "image")
+    """
+    if media_type != "auto":
+        return media_type
+
+    suffix = path.suffix.lower()
+
+    if suffix in EXTENSION_GROUPS["image"]:
+        return "image"
+    elif suffix in EXTENSION_GROUPS["audio"]:
+        return "audio"
+    elif suffix in EXTENSION_GROUPS["video"]:
+        return "video"
+    else:
+        return "image"
+
+
+def _get_mime_type(media_type: str, path: Path) -> str:
+    """获取文件的 MIME 类型
+
+    参数:
+        media_type: 媒体类型
+        path: 文件路径
+
+    返回:
+        MIME 类型字符串
+    """
+    suffix = path.suffix.lower()
+
+    if media_type in MIME_TYPES_MAP and suffix in MIME_TYPES_MAP[media_type]:
+        return MIME_TYPES_MAP[media_type][suffix]
+
+    return DEFAULT_MIMES.get(media_type, "application/octet-stream")

Undefined/skills/agents/file_analysis_agent/tools/cleanup_temp/config.json

@@ -0,0 +1,16 @@
+{
+    "type": "function",
+    "function": {
+        "name": "cleanup_temp",
+        "description": "清理指定的临时目录。用于在文件分析完成后删除临时下载的文件。",
+        "parameters": {
+            "type": "object",
+            "properties": {
+                "task_uuid": {
+                    "type": "string",
+                    "description": "任务 UUID（临时目录名），留空则清理所有 temp 目录"
+                }
+            }
+        }
+    }
+}

Undefined/skills/agents/file_analysis_agent/tools/cleanup_temp/handler.py

@@ -0,0 +1,35 @@
+from pathlib import Path
+from typing import Any, Dict
+import logging
+import shutil
+
+logger = logging.getLogger(__name__)
+
+
+async def execute(args: Dict[str, Any], context: Dict[str, Any]) -> str:
+    task_uuid: str = args.get("task_uuid", "")
+
+    temp_dir = Path.cwd() / "temp"
+
+    if not temp_dir.exists():
+        return "临时目录不存在"
+
+    try:
+        if task_uuid:
+            target_dir = temp_dir / task_uuid
+            if target_dir.exists():
+                shutil.rmtree(target_dir)
+                return f"已清理临时目录: {target_dir}"
+            else:
+                return f"临时目录不存在: {target_dir}"
+        else:
+            count = 0
+            for item in temp_dir.iterdir():
+                if item.is_dir():
+                    shutil.rmtree(item)
+                    count += 1
+            return f"已清理 {count} 个临时目录"
+
+    except Exception as e:
+        logger.exception(f"清理临时目录失败: {e}")
+        return f"清理临时目录失败: {e}"

Undefined/skills/agents/file_analysis_agent/tools/detect_file_type/config.json

@@ -0,0 +1,17 @@
+{
+    "type": "function",
+    "function": {
+        "name": "detect_file_type",
+        "description": "检测文件类型。Linux 系统使用 file 命令，非 Linux 系统使用魔数检测。",
+        "parameters": {
+            "type": "object",
+            "properties": {
+                "file_path": {
+                    "type": "string",
+                    "description": "本地文件路径"
+                }
+            },
+            "required": ["file_path"]
+        }
+    }
+}

Undefined/skills/agents/file_analysis_agent/tools/detect_file_type/handler.py

@@ -0,0 +1,221 @@
+import platform
+import subprocess
+from pathlib import Path
+from typing import Any, Dict
+import logging
+
+logger = logging.getLogger(__name__)
+
+MAGIC_NUMBERS: dict[bytes, str] = {
+    # 常用压缩格式
+    b"\x50\x4b\x03\x04": "ZIP Archive",
+    b"\x50\x4b\x05\x06": "ZIP Archive (empty)",
+    b"\x50\x4b\x07\x08": "ZIP Archive (spanned)",
+    # 文档格式
+    b"\x25\x50\x44\x46": "PDF Document",
+    b"\xd0\xcf\x11\xe0\xa1\xb1\x1a\xe1": "Microsoft Office Document (OLE)",
+    b"\x50\x4b\x03\x04\x14\x00\x06\x00": "OpenDocument Format",
+    # 图片格式
+    b"\xff\xd8\xff": "JPEG Image",
+    b"\x89\x50\x4e\x47\x0d\x0a\x1a\x0a": "PNG Image",
+    b"GIF8": "GIF Image",
+    b"\x49\x49\x2a\x00": "TIFF Image (little-endian)",
+    b"\x4d\x4d\x00\x2a": "TIFF Image (big-endian)",
+    b"\x42\x4d": "BMP Image",
+    # 音频格式
+    b"\x49\x44\x33": "MP3 Audio (ID3)",
+    b"\xff\xfb": "MP3 Audio (MPEG)",
+    b"\xff\xfa": "MP3 Audio (MPEG)",
+    b"\xff\xf3": "MP3 Audio (MPEG)",
+    b"\xff\xf2": "MP3 Audio (MPEG)",
+    # 视频格式
+    b"\x1a\x45\xdf\xa3": "WebM/Matroska Video",
+    b"\x00\x00\x00\x18ftypmp42": "MP4 Video (isom)",
+    b"\x00\x00\x00\x1cftypisom": "MP4 Video (isom)",
+    b"\x00\x00\x00\x20ftypmp41": "MP4 Video (isom)",
+    b"\x52\x49\x46\x46": "AVI/RIFF Video",
+    # 其他压缩格式
+    b"\x1f\x8b": "GZIP Archive",
+    b"\xfd7zXZ": "XZ Archive",
+    b"BZh": "BZIP2 Archive",
+}
+
+EXTENSION_MAP: dict[str, str] = {
+    ".pdf": "PDF Document",
+    ".docx": "Microsoft Word Document",
+    ".doc": "Microsoft Word Document (legacy)",
+    ".pptx": "Microsoft PowerPoint Document",
+    ".ppt": "Microsoft PowerPoint Document (legacy)",
+    ".xlsx": "Microsoft Excel Document",
+    ".xls": "Microsoft Excel Document (legacy)",
+    ".odt": "OpenDocument Text",
+    ".ods": "OpenDocument Spreadsheet",
+    ".odp": "OpenDocument Presentation",
+    ".txt": "Plain Text",
+    ".md": "Markdown Document",
+    ".rst": "reStructuredText Document",
+    ".log": "Log File",
+    ".json": "JSON Data",
+    ".yaml": "YAML Data",
+    ".yml": "YAML Data",
+    ".xml": "XML Data",
+    ".toml": "TOML Data",
+    ".ini": "INI Configuration",
+    ".py": "Python Source Code",
+    ".js": "JavaScript Source Code",
+    ".ts": "TypeScript Source Code",
+    ".c": "C Source Code",
+    ".cpp": "C++ Source Code",
+    ".cc": "C++ Source Code",
+    ".h": "C/C++ Header",
+    ".hpp": "C++ Header",
+    ".java": "Java Source Code",
+    ".go": "Go Source Code",
+    ".rs": "Rust Source Code",
+    ".php": "PHP Source Code",
+    ".rb": "Ruby Source Code",
+    ".swift": "Swift Source Code",
+    ".kt": "Kotlin Source Code",
+    ".sql": "SQL Script",
+    ".r": "R Source Code",
+    ".lua": "Lua Script",
+    ".sh": "Shell Script",
+    ".bash": "Bash Script",
+    ".html": "HTML Document",
+    ".htm": "HTML Document",
+    ".css": "CSS Stylesheet",
+    ".scss": "SCSS Stylesheet",
+    ".less": "LESS Stylesheet",
+    ".vue": "Vue Component",
+    ".jsx": "React Component",
+    ".tsx": "React TypeScript Component",
+    ".jpg": "JPEG Image",
+    ".jpeg": "JPEG Image",
+    ".png": "PNG Image",
+    ".gif": "GIF Image",
+    ".bmp": "BMP Image",
+    ".webp": "WebP Image",
+    ".svg": "SVG Image",
+    ".ico": "ICO Icon",
+    ".psd": "Photoshop Document",
+    ".mp3": "MP3 Audio",
+    ".wav": "WAV Audio",
+    ".flac": "FLAC Audio",
+    ".aac": "AAC Audio",
+    ".m4a": "MPEG-4 Audio",
+    ".ogg": "OGG Audio",
+    ".wma": "Windows Media Audio",
+    ".mp4": "MP4 Video",
+    ".avi": "AVI Video",
+    ".mov": "QuickTime Video",
+    ".mkv": "Matroska Video",
+    ".webm": "WebM Video",
+    ".flv": "Flash Video",
+    ".wmv": "Windows Media Video",
+    ".zip": "ZIP Archive",
+    ".tar": "TAR Archive",
+    ".gz": "GZIP Archive",
+    ".tgz": "GZIP-compressed TAR",
+    ".bz2": "BZIP2 Archive",
+    ".xz": "XZ Archive",
+    ".7z": "7-Zip Archive",
+    ".rar": "RAR Archive",
+    ".exe": "Windows Executable",
+    ".dll": "Windows Dynamic Library",
+    ".so": "Linux Shared Library",
+    ".dylib": "macOS Dynamic Library",
+    ".class": "Java Bytecode",
+    ".jar": "Java Archive",
+    ".war": "Java Web Archive",
+    ".apk": "Android Package",
+    ".ipa": "iOS App Package",
+    ".deb": "Debian Package",
+    ".rpm": "RPM Package",
+    ".dmg": "macOS Disk Image",
+    ".iso": "ISO Disk Image",
+    ".csv": "CSV Data",
+    ".tsv": "TSV Data",
+}
+
+
+async def execute(args: Dict[str, Any], context: Dict[str, Any]) -> str:
+    """执行文件类型检测
+
+    检测流程：
+    1. Linux 系统优先使用 `file` 命令
+    2. 其他系统或 file 命令失败时，使用文件头魔数 (Magic Number) 检测
+    3. 最后尝试使用文件扩展名检测
+    4. 如果都失败，返回 Unknown
+
+    参数:
+        args: 工具参数
+        context: 执行上下文
+
+    返回:
+        检测结果描述
+    """
+    file_path: str = args.get("file_path", "")
+    path = Path(file_path)
+
+    if not path.exists():
+        return f"错误：文件不存在 {file_path}"
+
+    if not path.is_file():
+        return f"错误：{file_path} 不是文件"
+
+    try:
+        file_size = path.stat().st_size
+        system = platform.system()
+
+        if system == "Linux":
+            result = _detect_by_file_command(path, file_size)
+            if result:
+                return result
+        else:
+            result = _detect_by_magic_number(path, file_size)
+            if result:
+                return result
+
+        result = _detect_by_extension(path)
+        if result:
+            return result
+
+        return f"Unknown file type (大小: {file_size} 字节)"
+
+    except Exception as e:
+        logger.exception(f"检测文件类型失败: {e}")
+        return f"检测文件类型失败: {e}"
+
+
+def _detect_by_file_command(path: Path, file_size: int) -> str | None:
+    try:
+        result = subprocess.run(
+            ["file", "-b", "-L", str(path)], capture_output=True, text=True, timeout=5
+        )
+        output = result.stdout.strip()
+        if output:
+            return f"{output} (大小: {file_size} 字节)"
+    except Exception as e:
+        logger.warning(f"file 命令执行失败: {e}")
+    return None
+
+
+def _detect_by_magic_number(path: Path, file_size: int) -> str | None:
+    try:
+        with open(path, "rb") as f:
+            header = f.read(32)
+
+        for magic, file_type in MAGIC_NUMBERS.items():
+            if header.startswith(magic):
+                return f"{file_type} (大小: {file_size} 字节)"
+    except Exception as e:
+        logger.warning(f"魔数检测失败: {e}")
+    return None
+
+
+def _detect_by_extension(path: Path) -> str | None:
+    ext = path.suffix.lower()
+    file_type = EXTENSION_MAP.get(ext)
+    if file_type:
+        return file_type
+    return None

Undefined/skills/agents/file_analysis_agent/tools/download_file/config.json

@@ -0,0 +1,21 @@
+{
+    "type": "function",
+    "function": {
+        "name": "download_file",
+        "description": "下载文件到临时目录。支持 URL 或 QQ 的 file_id。对于 URL 会先尝试获取文件大小，获取失败则拒绝下载。",
+        "parameters": {
+            "type": "object",
+            "properties": {
+                "file_source": {
+                    "type": "string",
+                    "description": "文件源：URL 或 QQ 的 file_id"
+                },
+                "max_size_mb": {
+                    "type": "number",
+                    "description": "最大允许的文件大小（MB），默认 100MB"
+                }
+            },
+            "required": ["file_source"]
+        }
+    }
+}

Undefined/skills/agents/file_analysis_agent/tools/download_file/handler.py

@@ -0,0 +1,124 @@
+import uuid
+from pathlib import Path
+from typing import Any, Dict
+import logging
+import httpx
+
+logger = logging.getLogger(__name__)
+
+SIZE_LIMITS = {
+    "text": 10 * 1024 * 1024,
+    "code": 5 * 1024 * 1024,
+    "pdf": 50 * 1024 * 1024,
+    "docx": 20 * 1024 * 1024,
+    "pptx": 20 * 1024 * 1024,
+    "xlsx": 10 * 1024 * 1024,
+    "image": 10 * 1024 * 1024,
+    "audio": 50 * 1024 * 1024,
+    "video": 100 * 1024 * 1024,
+    "archive": 100 * 1024 * 1024,
+}
+
+DEFAULT_SIZE_LIMIT = 100 * 1024 * 1024
+
+
+async def execute(args: Dict[str, Any], context: Dict[str, Any]) -> str:
+    file_source: str = args.get("file_source", "")
+    max_size_mb: float = args.get("max_size_mb", 100)
+
+    if not file_source:
+        return "错误：文件源不能为空"
+
+    task_uuid: str = uuid.uuid4().hex[:16]
+    temp_dir: Path = Path.cwd() / "temp" / task_uuid
+    temp_dir.mkdir(parents=True, exist_ok=True)
+
+    is_url: bool = file_source.startswith("http://") or file_source.startswith(
+        "https://"
+    )
+
+    if is_url:
+        return await _download_from_url(file_source, temp_dir, max_size_mb, task_uuid)
+    else:
+        return await _download_from_file_id(file_source, temp_dir, context, task_uuid)
+
+
+async def _download_from_url(
+    url: str, temp_dir: Path, max_size_mb: float, task_uuid: str
+) -> str:
+    max_size_bytes: int = int(max_size_mb * 1024 * 1024)
+
+    async with httpx.AsyncClient(timeout=60.0) as client:
+        try:
+            logger.info(f"正在获取文件大小: {url}")
+            head_response = await client.head(url, timeout=30.0)
+            content_length = head_response.headers.get("content-length")
+
+            if content_length is None:
+                return "错误：无法获取文件大小，拒绝下载"
+
+            file_size = int(content_length)
+            if file_size > max_size_bytes:
+                return f"错误：文件大小 ({file_size / 1024 / 1024:.2f}MB) 超过限制 ({max_size_mb}MB)"
+
+            logger.info(f"文件大小: {file_size / 1024 / 1024:.2f}MB，允许下载")
+
+            logger.info("正在下载文件...")
+            response = await client.get(url, timeout=120.0)
+            response.raise_for_status()
+
+            filename = _extract_filename_from_url(url)
+            if not filename or "." not in filename:
+                filename = f"downloaded_{task_uuid}"
+
+            file_path = temp_dir / filename
+            file_path.write_bytes(response.content)
+
+            logger.info(f"文件已保存到: {file_path}")
+            return str(file_path)
+
+        except httpx.TimeoutException:
+            return "错误：下载超时"
+        except httpx.HTTPStatusError as e:
+            return f"错误：HTTP 错误 {e.response.status_code}"
+        except Exception as e:
+            logger.exception(f"下载失败: {e}")
+            return f"错误：下载失败 - {e}"
+
+
+async def _download_from_file_id(
+    file_id: str, temp_dir: Path, context: Dict[str, Any], task_uuid: str
+) -> str:
+    get_image_url_callback = context.get("get_image_url_callback")
+    if not get_image_url_callback:
+        return "错误：file_id 模式需要 get_image_url_callback"
+
+    try:
+        logger.info(f"正在解析 file_id: {file_id}")
+        url = await get_image_url_callback(file_id)
+        if not url:
+            return f"错误：无法将 file_id {file_id} 解析为 URL"
+
+        logger.info(f"获取到 URL: {url}")
+
+        async with httpx.AsyncClient(timeout=120.0) as client:
+            response = await client.get(url, timeout=120.0)
+            response.raise_for_status()
+
+            filename = f"file_{file_id}"
+            file_path = temp_dir / filename
+            file_path.write_bytes(response.content)
+
+            logger.info(f"文件已保存到: {file_path}")
+            return str(file_path)
+
+    except Exception as e:
+        logger.exception(f"下载失败（file_id 模式）: {e}")
+        return f"错误：下载失败 - {e}"
+
+
+def _extract_filename_from_url(url: str) -> str:
+    if "?" in url:
+        url = url.split("?")[0]
+    filename = url.split("/")[-1]
+    return filename

Undefined/skills/agents/file_analysis_agent/tools/extract_archive/config.json

@@ -0,0 +1,25 @@
+{
+    "type": "function",
+    "function": {
+        "name": "extract_archive",
+        "description": "提取压缩包文件内容。支持 zip, tar, gz, bz2, xz, 7z, rar 等格式。支持列出文件列表或解压到临时目录。",
+        "parameters": {
+            "type": "object",
+            "properties": {
+                "file_path": {
+                    "type": "string",
+                    "description": "本地压缩包路径"
+                },
+                "action": {
+                    "type": "string",
+                    "description": "操作类型：'list'（列出文件列表）或 'extract'（解压到临时目录），默认为 'list'"
+                },
+                "extract_path": {
+                    "type": "string",
+                    "description": "解压目标路径（可选），默认为临时目录"
+                }
+            },
+            "required": ["file_path"]
+        }
+    }
+}