Communitygithub.com

film-crew

导演身边的剧组:拍片的方法(film-director 里的 video-film-maker、blender-previz)管怎么拍,这里是拍的时候用的四样工具,装一次全有。backends 把任务单(一次生成要什么的 JSON)交给视频、图像模型的官方接口,出片出图,key 走环境变量,每个模型一张能力卡;viewer 把拍片的运行目录做成网页——拍片进行中的进度页(每一次生成都能看,花了多少时间和钱)、拍完的展示台(每一镜喂进模型的图、首尾帧、参考图、prompt、参数和成片对照)和节点图(每镜谁喂给了谁,能导出 ComfyUI 工作流);comfyui 把每一镜放进 ComfyUI 给人看、给人接手,只回放不生成;formats 是它们之间两份交接文件(任务单、镜头清单)的格式、检查脚本,和把运行目录读成清单的脚本。Use when a film run needs a video or an image generated from a job sheet, a look at the exact request before it is sent, a model chosen by its capability card (feeds, length, reference images, price, speed), a live progress page while shots are generating, a showcase page or node graph of what each shot was fed and what came out, the shots opened in ComfyUI as workflows, or a run directory turned into a shots manifest — 任务单 / 出片 / 出图 / 选模型 / 能力卡 / 进度页 / 每一次生成都能看 / 展示台 / 节点图 / 镜头清单 / 读运行目录 / 接 ComfyUI / 导出 ComfyUI / 百炼 / 火山方舟 / MiniMax / OpenAI Images / film-crew.

film-crew 是什么?

film-crew is a Codex agent skill that 导演身边的剧组:拍片的方法(film-director 里的 video-film-maker、blender-previz)管怎么拍,这里是拍的时候用的四样工具,装一次全有。backends 把任务单(一次生成要什么的 JSON)交给视频、图像模型的官方接口,出片出图,key 走环境变量,每个模型一张能力卡;viewer 把拍片的运行目录做成网页——拍片进行中的进度页(每一次生成都能看,花了多少时间和钱)、拍完的展示台(每一镜喂进模型的图、首尾帧、参考图、prompt、参数和成片对照)和节点图(每镜谁喂给了谁,能导出 ComfyUI 工作流);comfyui 把每一镜放进 ComfyUI 给人看、给人接手,只回放不生成;formats 是它们之间两份交接文件(任务单、镜头清单)的格式、检查脚本,和把运行目录读成清单的脚本。Use when a film run needs a video or an image generated from a job sheet, a look at the exact request before it is sent, a model chosen by its capability card (feeds, length, reference images, price, speed), a live progress page while shots are generating, a showcase page or node graph of what each shot was fed and what came out, the shots opened in ComfyUI as workflows, or a run directory turned into a shots manifest — 任务单 / 出片 / 出图 / 选模型 / 能力卡 / 进度页 / 每一次生成都能看 / 展示台 / 节点图 / 镜头清单 / 读运行目录 / 接 ComfyUI / 导出 ComfyUI / 百炼 / 火山方舟 / MiniMax / OpenAI Images / film-crew.

兼容平台~Claude Code✓Codex CLI~Cursor
npx skills add https://github.com/memoryoffish/film-director-skill-plugin/tree/HEAD/plugins/film-crew

在你喜欢的 AI 中提问

打开一个已预加载此 Agent Skill 的新对话。

文档

film-crew 是做什么的?

In a nutshell: the filming methods determine "how to shoot"; here are the tools used during the shoot—select models, watch progress, present the finished work. Four parallel modules, each with its own overview.md; all paths are relative to this directory.

What each of the four modules does

ModuleWhat it doesWhen to useRead first
backends/Job sheets in, videos and images outTo generate a video or a set of images; to preview what request will be sent; to select a modelbackends/overview.md
viewer/Run directory → web pages for humansDuring a film run: watch each generation (progress page); after the run: see what each shot consumed and produced (showcase, node graph)viewer/overview.md
comfyui/Each shot placed into ComfyUIWhen a human wants to see in ComfyUI how a shot was built, and then reuse the key frames and outputscomfyui/overview.md
formats/Format specifications for two interchange filesTo write or validate a job sheet; to read a run directory into a shots manifestformats/overview.md

The four modules communicate only through two files: job sheet (specifies what to generate in one request, handed to the backend) and shots manifest (documents what each shot in a film used and produced, handed to the two viewers).

Where they appear during a film shoot

  1. Start the shoot: Launch the progress page inside a tmux session; every subsequent generation will be listed there—python3 viewer/scripts/progress.py <output directory> --jobs <project>/clips --prices backends/cards/prices.json --every 60.
  2. Generate: Each generation writes a job sheet into <project>/jobs/ (for video) or <project>/jobs_img/ (for images). First, python3 backends/scripts/filmgen.py request <sheet> to preview what will be sent, then run. The backends call the official APIs of each provider: Alibaba Cloud Bailian (Wan 3.0, HappyHorse 1.1), Volcano Engine Ark (Seedance 2.5 / 2.0), MiniMax (H3-Max), OpenAI Images (gpt-image 2.6). Which model to use is chosen via backends/cards/.
  3. Finish: Write a shot table meta.json in the run directory (each shot's name, its start and end timestamps in the finished film, which generation record was used), then run python3 formats/read_run.py <config> --out <manifest directory> to produce a shots manifest, and python3 formats/check_manifest.py to validate it.
  4. Present to the viewer: The web page is produced by python3 viewer/scripts/console.py <manifest…> <output directory>; to open in ComfyUI, run python3 comfyui/scripts/shots2comfy.py <manifest…> --comfy <ComfyUI directory>. The node graph on the web page follows the same workflow structure as ComfyUI; to include it, first run shots2comfy.py, then add --workflows to console.py.

Requirements: Python 3.10+. backends, formats, the progress page, and shots2comfy.py rely only on the standard library (backends uses ffprobe to measure output video duration); the showcase (console.py) additionally needs Pillow and ffmpeg; send requires SSH with tmux on the remote machine; ComfyUI node packages run within ComfyUI's own environment (using its bundled Pillow and numpy).

For the parameters of each step and what to check after each step, see that module's overview.md.

Rules

  • Keys are read from environment variables only (DASHSCOPE_API_KEY, ARK_API_KEY, MINIMAX_API_KEY, OPENAI_API_KEY); never written into any file, never printed. request works even without a key set.
  • Viewers are read-only: they do not modify the run directory and do not call generation APIs. If a human sees adjustments needed in ComfyUI or on the web page, they tell the agent, who makes changes and submits a new job.
  • Preview before sending: For every job sheet, run request first. To redo or change parameters, save the modified sheet under a new name.
  • Preview the generated page yourself before handing it over to the human.

相关技能