@remixmate/cli 0.9.72 → 0.9.74

package/skills/video-parser/scripts/deconstruct_video.py

@@ -1,21 +1,21 @@
 #!/usr/bin/env python3
-"""
-视频解构脚本 - 将视频拆解为可复用的内容资产
-自动提取：音频、ASR 文本（含逐字时间戳）、场景切分、关键帧图片。
-输出结构化资产清单 JSON。
 
-依赖：本地须安装 ffmpeg（brew install ffmpeg）
 
-用法:
-  python deconstruct_video.py --url "https://example.com/video.mp4"
-  python deconstruct_video.py --local /path/to/video.mp4
-  python deconstruct_video.py --url "https://example.com/video.mp4" --skip-asr
-  python deconstruct_video.py --url "https://example.com/video.mp4" --json-output
 
-环境变量:
-  MM_API_BASE_URL  - 后端 API 地址（默认: https://api-agent.remixmate.ai/api）
-  PRIV_TOKEN       - PrivToken 认证令牌
-"""
+
+
+
+
+
+
+
+
+
+
+
+
+
+
 
 import argparse
 import json
@@ -29,27 +29,27 @@ import urllib.request
 
 API_BASE_URL = os.environ.get("MM_API_BASE_URL", "https://api-agent.remixmate.ai/api")
 
-PRIVATE_TOKEN = ""  # 在 main() 中通过 resolve_token() 初始化
+PRIVATE_TOKEN = ""
 SKILL_NAME = "video-parser"
 AGENT_NAME = os.environ.get("AGENT_NAME", "")
 
-# 是否静默模式（--json-output 时抑制进度输出）
+
 QUIET = False
 
 
 def log(msg: str) -> None:
-    """打印进度信息（静默模式下不输出）"""
+
     if not QUIET:
         print(msg, flush=True)
 
 
 def log_err(msg: str) -> None:
-    """打印错误信息到 stderr（始终输出）"""
+
     print(msg, file=sys.stderr, flush=True)
 
 
 def resolve_token(cli_token: str = None) -> str:
-    """优先从 CLI 参数读取 token，其次环境变量，最后报错退出"""
+
     if cli_token:
         return cli_token
     token = os.environ.get("PRIV_TOKEN", "")
@@ -63,7 +63,7 @@ def resolve_token(cli_token: str = None) -> str:
 
 
 def _build_headers() -> dict:
-    """构建统一请求头（X-Priv-Token 认证头）"""
+
     headers = {
         "Content-Type": "application/json",
         "X-Priv-Token": PRIVATE_TOKEN,
@@ -74,12 +74,12 @@ def _build_headers() -> dict:
     return headers
 
 
-# ---------------------------------------------------------------------------
-# 前置检查
-# ---------------------------------------------------------------------------
+
+
+
 
 def check_ffmpeg() -> None:
-    """检查 ffmpeg 是否已安装"""
+
     try:
         subprocess.run(
             ["ffmpeg", "-version"],
@@ -92,13 +92,13 @@ def check_ffmpeg() -> None:
         sys.exit(1)
 
 
-# ---------------------------------------------------------------------------
-# 步骤 1：获取视频到本地
-# ---------------------------------------------------------------------------
+
+
+
 
 def download_video(url: str, output_dir: str) -> str:
-    """下载远程视频到本地，返回本地文件路径"""
-    # 从 URL 推断文件扩展名
+
+
     url_path = url.split("?")[0]
     ext = os.path.splitext(url_path)[1] or ".mp4"
     local_path = os.path.join(output_dir, f"source_video{ext}")
@@ -109,7 +109,7 @@ def download_video(url: str, output_dir: str) -> str:
         with urllib.request.urlopen(req, timeout=300) as resp:
             with open(local_path, "wb") as f:
                 while True:
-                    chunk = resp.read(1024 * 1024)  # 1MB chunks
+                    chunk = resp.read(1024 * 1024)
                     if not chunk:
                         break
                     f.write(chunk)
@@ -123,7 +123,7 @@ def download_video(url: str, output_dir: str) -> str:
 
 
 def get_video_duration_ms(video_path: str) -> int:
-    """使用 ffprobe 获取视频时长（毫秒）"""
+
     try:
         result = subprocess.run(
             [
@@ -140,12 +140,12 @@ def get_video_duration_ms(video_path: str) -> int:
         return 0
 
 
-# ---------------------------------------------------------------------------
-# 步骤 2：提取音频
-# ---------------------------------------------------------------------------
+
+
+
 
 def extract_audio(video_path: str, output_dir: str) -> str:
-    """从视频中提取音频为 MP3，返回音频文件路径"""
+
     audio_path = os.path.join(output_dir, "audio.mp3")
     log("[2/6] 正在提取音频...")
 
@@ -169,12 +169,12 @@ def extract_audio(video_path: str, output_dir: str) -> str:
     return audio_path
 
 
-# ---------------------------------------------------------------------------
-# 步骤 3：上传音频到 OSS
-# ---------------------------------------------------------------------------
+
+
+
 
 def upload_audio(audio_path: str) -> str:
-    """上传音频文件到 OSS（POST /file/uploadMediaFile，multipart），返回文件 URL"""
+
     log("[3/6] 正在上传音频到 OSS...")
 
     url = f"{API_BASE_URL}/file/uploadMediaFile"
@@ -184,27 +184,27 @@ def upload_audio(audio_path: str) -> str:
     with open(audio_path, "rb") as f:
         file_data = f.read()
 
-    # 构建 multipart/form-data body
+
     body = b""
-    # file 字段
+
     body += f"--{boundary}\r\n".encode()
     body += f'Content-Disposition: form-data; name="file"; filename="{filename}"\r\n'.encode()
     body += b"Content-Type: audio/mpeg\r\n\r\n"
     body += file_data
     body += b"\r\n"
-    # title 字段
+
     body += f"--{boundary}\r\n".encode()
     body += b'Content-Disposition: form-data; name="title"\r\n\r\n'
     body += filename.encode()
     body += b"\r\n"
-    # role / retention：ab-api 落素材表时必填角色。抽离音频只为 ASR，属中间产物；
-    # 源视频是第三方的，检索侧据 role 默认排除（license=third_party_unknown）。
+
+
     for name, value in (("role", "source_audio"), ("retention", "intermediate")):
         body += f"--{boundary}\r\n".encode()
         body += f'Content-Disposition: form-data; name="{name}"\r\n\r\n'.encode()
         body += value.encode()
         body += b"\r\n"
-    # 结束
+
     body += f"--{boundary}--\r\n".encode()
 
     headers = {
@@ -244,12 +244,12 @@ def upload_audio(audio_path: str) -> str:
     return file_url
 
 
-# ---------------------------------------------------------------------------
-# 步骤 4：ASR 语音识别
-# ---------------------------------------------------------------------------
+
+
+
 
 def run_asr(audio_url: str) -> dict:
-    """调用 ASR 接口（POST /tool/asr），返回识别结果"""
+
     log("[4/6] 正在进行语音识别 (ASR)...")
 
     url = f"{API_BASE_URL}/tool/asr"
@@ -289,15 +289,15 @@ def run_asr(audio_url: str) -> dict:
     }
 
 
-# ---------------------------------------------------------------------------
-# 步骤 5：场景切变检测 & 关键帧提取
-# ---------------------------------------------------------------------------
+
+
+
 
 def detect_scenes(video_path: str, threshold: float) -> list:
-    """
-    使用 ffmpeg select 滤镜检测场景切变，返回切点时间戳列表（秒）。
-    始终包含 0.0 作为第一个切点。
-    """
+
+
+
+
     log(f"[5/6] 正在检测场景切变 (阈值={threshold})...")
 
     try:
@@ -310,7 +310,7 @@ def detect_scenes(video_path: str, threshold: float) -> list:
             ],
             capture_output=True, text=True,
         )
-        # 从 stderr 解析 pts_time
+
         output = result.stderr
     except Exception as e:
         log_err(f"场景检测失败: {e}")
@@ -321,7 +321,7 @@ def detect_scenes(video_path: str, threshold: float) -> list:
         ts = float(match.group(1))
         timestamps.append(ts)
 
-    # 去重并排序，始终包含 0.0
+
     timestamps = sorted(set(timestamps))
     if not timestamps or timestamps[0] != 0.0:
         timestamps.insert(0, 0.0)
@@ -331,7 +331,7 @@ def detect_scenes(video_path: str, threshold: float) -> list:
 
 
 def extract_keyframes(video_path: str, timestamps: list, output_dir: str) -> list:
-    """按时间戳列表从视频截取关键帧，返回关键帧信息列表"""
+
     keyframes_dir = os.path.join(output_dir, "keyframes")
     os.makedirs(keyframes_dir, exist_ok=True)
 
@@ -368,12 +368,12 @@ def extract_keyframes(video_path: str, timestamps: list, output_dir: str) -> lis
     return keyframes
 
 
-# ---------------------------------------------------------------------------
-# 步骤 6：汇总输出
-# ---------------------------------------------------------------------------
+
+
+
 
 def build_scenes(timestamps: list, video_duration_sec: float, keyframes: list) -> list:
-    """根据切点时间戳构建场景分段列表"""
+
     scenes = []
     keyframe_map = {kf["index"]: kf["localPath"] for kf in keyframes}
 
@@ -399,7 +399,7 @@ def build_result(
     scenes: list,
     keyframes: list,
 ) -> dict:
-    """汇总所有数据为最终资产清单"""
+
     result = {
         "source": {
             "url": source_url or None,
@@ -417,9 +417,9 @@ def build_result(
     return result
 
 
-# ---------------------------------------------------------------------------
-# Main
-# ---------------------------------------------------------------------------
+
+
+
 
 def main():
     parser = argparse.ArgumentParser(
@@ -492,7 +492,7 @@ Examples:
 
     args = parser.parse_args()
 
-    # 校验输入
+
     if not args.url and not args.local:
         log_err("请提供 --url 或 --local 参数指定视频来源")
         parser.print_help()
@@ -507,29 +507,29 @@ Examples:
         log_err("--scene-threshold 范围为 0.0-1.0")
         sys.exit(1)
 
-    # 静默模式
+
     global QUIET
     QUIET = args.json_output
 
-    # 初始化 token（如果需要 ASR，则必须有 token）
+
     global PRIVATE_TOKEN
     if not args.skip_asr:
         PRIVATE_TOKEN = resolve_token(args.priv_token)
     else:
-        # 即使跳过 ASR，如果提供了 token 也保存（可能用于其他扩展）
+
         PRIVATE_TOKEN = args.priv_token or os.environ.get("PRIV_TOKEN", "")
 
-    # 前置检查
+
     check_ffmpeg()
 
-    # 输出目录
+
     output_dir = args.output_dir or f"./deconstructed_{int(time.time())}"
     os.makedirs(output_dir, exist_ok=True)
     log(f"输出目录: {os.path.abspath(output_dir)}")
 
-    # -----------------------------------------------------------------------
-    # [1] 获取视频到本地
-    # -----------------------------------------------------------------------
+
+
+
     source_url = args.url or ""
     if args.url:
         video_path = download_video(args.url, output_dir)
@@ -537,19 +537,19 @@ Examples:
         video_path = os.path.abspath(args.local)
         log(f"[1/6] 使用本地视频: {video_path}")
 
-    # 获取视频时长
+
     video_duration_ms = get_video_duration_ms(video_path)
     video_duration_sec = video_duration_ms / 1000.0 if video_duration_ms > 0 else 0.0
     log(f"   视频时长: {video_duration_sec:.1f}s ({video_duration_ms}ms)")
 
-    # -----------------------------------------------------------------------
-    # [2] 提取音频
-    # -----------------------------------------------------------------------
+
+
+
     audio_path = extract_audio(video_path, output_dir)
 
-    # -----------------------------------------------------------------------
-    # [3-4] 上传音频 + ASR
-    # -----------------------------------------------------------------------
+
+
+
     audio_oss_url = ""
     asr_result = {"text": "", "duration": 0, "utterances": []}
 
@@ -560,9 +560,9 @@ Examples:
         log("[3/6] 跳过音频上传（--skip-asr）")
         log("[4/6] 跳过 ASR（--skip-asr）")
 
-    # -----------------------------------------------------------------------
-    # [5-6] 场景检测 + 关键帧提取
-    # -----------------------------------------------------------------------
+
+
+
     scenes = []
     keyframes = []
 
@@ -574,9 +574,9 @@ Examples:
         log("[5/6] 跳过场景检测（--skip-keyframes）")
         log("[6/6] 跳过关键帧提取（--skip-keyframes）")
 
-    # -----------------------------------------------------------------------
-    # 汇总输出
-    # -----------------------------------------------------------------------
+
+
+
     final = build_result(
         source_url=source_url,
         local_video_path=video_path,
@@ -588,13 +588,13 @@ Examples:
         keyframes=keyframes,
     )
 
-    # 写入 JSON
+
     json_path = os.path.join(output_dir, "deconstruction.json")
     with open(json_path, "w", encoding="utf-8") as f:
         json.dump(final, f, ensure_ascii=False, indent=2)
 
     if args.json_output:
-        # 管道模式：仅输出 JSON 到 stdout
+
         print(json.dumps(final, ensure_ascii=False, indent=2))
     else:
         log(f"\n资产清单已保存: {json_path}")

package/skills/video-parser/scripts/parse_via_render.py

@@ -1,18 +1,18 @@
 #!/usr/bin/env python3
-"""
-视频解构脚本 — 通过 ab-render HTTP 接口实现，无需本地 ffmpeg。
 
-调用链：
-  POST $RENDER_API_URL/parse        → { taskId }
-  POST $RENDER_API_URL/parseStatus  → { status, progress, result?, error? }
 
-必要环境变量（由 ab-agent 自动注入）：
-  RENDER_API_URL   ab-render 服务地址（默认 https://api-render.remixmate.ai；指向本地/staging 时覆盖）
-  PRIV_TOKEN       ab-api 私有 token
 
-可选环境变量：
-  CONVERSATION_ID  对话 ID，用于文件关联
-"""
+
+
+
+
+
+
+
+
+
+
+
 
 import argparse
 import json
@@ -22,7 +22,7 @@ import time
 import urllib.error
 import urllib.request
 
-# ─── 环境变量 ─────────────────────────────────────────────────────────────────
+
 
 RENDER_API_URL = (
     os.environ.get("RENDER_API_URL")
@@ -32,10 +32,10 @@ RENDER_API_URL = (
 PRIV_TOKEN = os.environ.get("PRIV_TOKEN", "")
 CONVERSATION_ID = os.environ.get("CONVERSATION_ID", "")
 
-# ─── HTTP 工具 ────────────────────────────────────────────────────────────────
+
 
 def api_post(path: str, body: dict, timeout: int = 30) -> dict:
-    """向 ab-render 发送 POST 请求，返回解析后的 JSON。"""
+
     url = f"{RENDER_API_URL}{path}"
     data = json.dumps(body).encode()
 
@@ -60,7 +60,7 @@ def log(msg: str, silent: bool = False) -> None:
         print(msg, flush=True)
 
 
-# ─── 主逻辑 ───────────────────────────────────────────────────────────────────
+
 
 def main() -> None:
     parser = argparse.ArgumentParser(
@@ -79,7 +79,7 @@ def main() -> None:
     )
     args = parser.parse_args()
 
-    # ── Pre-flight check ─────────────────────────────────────────────────────
+
     if not RENDER_API_URL:
         print(
             "❌ RENDER_API_URL env var is not configured.\n"
@@ -88,7 +88,7 @@ def main() -> None:
         )
         sys.exit(1)
 
-    # ── 1. Submit the deconstruction task ───────────────────────────────────
+
     log(f"[video-parser] Submitting deconstruction task: {args.url[:80]}", args.json_output)
 
     resp = api_post("/parse", {
@@ -105,9 +105,9 @@ def main() -> None:
     task_id: str = resp["data"]["taskId"]
     log(f"[video-parser] task accepted, taskId={task_id}", args.json_output)
 
-    # ── 2. Poll status until completion ─────────────────────────────────────
-    max_wait_sec = 600   # Wait up to 10 minutes
-    poll_interval = 3    # Poll every 3 seconds
+
+    max_wait_sec = 600
+    poll_interval = 3
     elapsed = 0
 
     while elapsed < max_wait_sec:
@@ -116,7 +116,7 @@ def main() -> None:
 
         try:
             status_resp = api_post("/parseStatus", {"taskId": task_id})
-        except Exception as e:  # noqa: BLE001
+        except Exception as e:
             log(f"[video-parser] polling error (will retry): {e}", args.json_output)
             continue
 

package/skills/web-screenshot/scripts/_media_screenshot/__init__.py

@@ -1,8 +1,8 @@
-"""media-screenshot internal package.
 
-Public surface used by the entry-point scripts. Anything else under this
-package is implementation detail.
-"""
+
+
+
+
 from __future__ import annotations
 
 from . import cli_args, scenes, template, trim

package/skills/web-screenshot/scripts/_media_screenshot/bootstrap.py

@@ -1,4 +1,4 @@
-"""Auto-install playwright + chromium on first run."""
+
 from __future__ import annotations
 
 import subprocess
@@ -14,7 +14,7 @@ def _pip_install(pkg: str) -> None:
 
 def _ensure_playwright_pkg() -> None:
     try:
-        import playwright.sync_api  # noqa: F401
+        import playwright.sync_api
         return
     except ImportError:
         pass
@@ -22,11 +22,11 @@ def _ensure_playwright_pkg() -> None:
 
 
 def ensure_browser(name: str = "chromium") -> None:
-    """Install the given browser engine via `playwright install <name>`.
 
-    Must install the engine that is about to be launched: installing chromium
-    and then relaunching firefox fails the same way a second time.
-    """
+
+
+
+
     print(f"[bootstrap] playwright install {name} ...", file=sys.stderr)
     subprocess.check_call(
         [sys.executable, "-m", "playwright", "install", name]
@@ -34,5 +34,5 @@ def ensure_browser(name: str = "chromium") -> None:
 
 
 def ensure_runtime() -> None:
-    """Make sure the playwright pip package is importable."""
+
     _ensure_playwright_pkg()

package/skills/web-screenshot/scripts/_media_screenshot/browser.py

@@ -1,4 +1,4 @@
-"""Browser context / locator / wait helpers."""
+
 from __future__ import annotations
 
 from typing import Any
@@ -49,14 +49,14 @@ def _is_missing_executable(e: Exception) -> bool:
 
 def launch_with_browser_install(p, browser_name: str | None, launch_kwargs: dict | None = None):
     btype = pick_browser_type(p, browser_name)
-    engine = btype.name  # "chromium" | "firefox" | "webkit"
+    engine = btype.name
     try:
         return btype.launch(**(launch_kwargs or {}))
     except Exception as e:
         if not _is_missing_executable(e):
             raise
-    # 引擎没装：装「这次要启动的那个」引擎再试一次。装不上 / 装了仍起不来时给一句
-    # 能据以行动的话，而不是 30 行 Playwright traceback。
+
+
     hint = (
         f"浏览器引擎 {engine} 在当前环境不可用。"
         + ("去掉 browser 参数改用默认的 chromium 即可。" if engine != "chromium"

package/skills/web-screenshot/scripts/_media_screenshot/cli_args.py

@@ -1,4 +1,4 @@
-"""Argparse helpers shared by ``screenshot.py`` and ``record.py``."""
+
 from __future__ import annotations
 
 import argparse
@@ -9,7 +9,7 @@ from pathlib import Path
 
 
 def add_common_args(ap: argparse.ArgumentParser) -> None:
-    """Add the args common to both screenshot and record entry points."""
+
     ap.add_argument(
         "-b", "--browser", choices=["chromium", "firefox", "webkit"],
         default="chromium", help="浏览器内核（默认 chromium）",
@@ -59,7 +59,7 @@ def parse_cookies(raw: str) -> list[dict]:
 
 
 def build_storage(args) -> tuple[str | None, bool]:
-    """Returns (storage_state_path, is_temp). Caller must unlink if is_temp."""
+
     if getattr(args, "storage_state", None):
         return args.storage_state, False
     if getattr(args, "cookies", None):
@@ -92,7 +92,7 @@ def parse_clip(raw: str) -> list[float]:
 
 
 def common_args_to_cfg(args, cfg: dict) -> None:
-    """Merge the common-args values into a cfg dict (shared by entry points)."""
+
     if args.device:
         cfg["device"] = args.device
     if args.viewport:

package/skills/web-screenshot/scripts/_media_screenshot/js_loader.py

@@ -1,4 +1,4 @@
-"""Load .js files from `js/` as strings, cached at import time."""
+
 from __future__ import annotations
 
 from functools import lru_cache
@@ -9,6 +9,6 @@ _JS_DIR = Path(__file__).resolve().parent / "js"
 
 @lru_cache(maxsize=None)
 def load(name: str) -> str:
-    """Return the contents of js/<name>.js. Raises FileNotFoundError if missing."""
+
     path = _JS_DIR / f"{name}.js"
     return path.read_text(encoding="utf-8")

package/skills/web-screenshot/scripts/_media_screenshot/overlay.py

@@ -1,4 +1,4 @@
-"""Small wrappers around the overlay-management JS snippets."""
+
 from __future__ import annotations
 
 from . import js_loader