Turn 知易 knowledge cards into scripts for batch video production with MoneyPrinterTurbo. The workflow also transcribes video audio locally with faster-whisper and imports the text as cards for future scripts.
知易知识库 → 短视频 流水线。把 knowledge_cards（蓝/绿/红卡）批量生成书摘/知识类短视频（复用 MoneyPrinterTurbo），并提供本地 faster-whisper 转写桥，把任意视频转文字入库成为脚本源。触发：用户想"把书/知识库做成视频""批量产出书摘视频""把视频转写进知识库"。
Run score Not tested
See how to installReview the instructions the agent will receive before you install.
knowledge_cards 直接变成可渲染的短视频脚本，推给本地 MoneyPrinterTurbo (MPT) 批量生成书摘/知识类视频（口播 + 字幕 + 素材 + BGM）。kb_to_mpt.py — 读知识库卡片 → 组装 video_script + video_terms → POST /api/v1/videos 推 MPT。transcribe_fw.py — faster-whisper（项目机器人同款 STT）转写音频 → .vtt/.txt，供入库。build_naval_cards.py（范例：书摘视频转写 → 建专题蓝卡）与 ingest_video_material.py（vtt → 蓝卡通用入库）。| 依赖 | 说明 | 准备 |
|---|---|---|
| MoneyPrinterTurbo | 视频生成引擎（C:/E/MoneyPrinterTurbo-main，默认 :8080） | 真机跑 webui.bat 启动 |
| ffmpeg | 音频解码（whisper 需要） | winget install ffmpeg，或复用知易内置 _bin/ffmpeg.exe |
| faster-whisper | 转写引擎（CT2 格式权重） | 见下方"权重准备" |
| 知易 DB | C:/D/zhiyi/backend/data/antinet.db | 已有 |
权重准备（一次性，绕开 huggingface_hub 沙箱限制，直接目录加载）：
把 Systran/faster-whisper-<size> 仓库的 model.bin + config.json + tokenizer.json + vocabulary.txt
放到 C:/D/zhiyi/data/whisper_models/<size>/，transcribe_fw.py 即以目录路径直接加载，不再触发任何下载。
推荐 small（中文可用，~461MB）；base 更快但中文弱。
# 1) 某专题全部卡片 → 一条视频（dry-run 先看文案）
python kb_to_mpt.py --project-id <PID> --dry-run
# 2) 按关键词搜卡片生成（如"纳瓦尔宝典"）
python kb_to_mpt.py --search "纳瓦尔" --subject "纳瓦尔宝典精读" --dry-run
# 3) 真机启 MPT 后实际上传生成
python kb_to_mpt.py --project-id <PID> --mpt-url http://127.0.0.1:8080
MPT 收到 video_subject + video_script（卡片拼装的连贯口播）+ video_terms（标题抽词作素材关键词），自动产出高清短视频。
# 1) 下载音频（用 yt-dlp 技能；B站 bestaudio 无需 ffmpeg）
python yt-dlp-video-downloader/scripts/run_yt_dlp.py -f bestaudio \
-o "material/audio/%(id)s.%(ext)s" "<视频URL>"
# 2) faster-whisper 本地转写
python transcribe_fw.py "material/audio/<BV>.m4a" --out material/audio --model small --lang zh
# 3) 切片成蓝卡入库（通用 vtt→卡）
python ingest_video_material.py material/audio --url "<视频URL>" --project-id <PID>
# 或：书摘类按段落建专题（范例脚本）
python build_naval_cards.py material/audio/<BV>.txt
书/视频 → [yt-dlp 下载音频] → [transcribe_fw 转写] → [build_* 建蓝卡专题]
→ [kb_to_mpt 组装脚本] → [MPT 渲染成片] → 发布分发(ai-trend-publish)
循环复用：换一本书/一个专题，只改 --project-id 或 --search，即可批量产出下一支视频。
kb_to_mpt.py 的 DB 路径硬编码为 C:/D/zhiyi/backend/data/antinet.db，知易项目用户即开即用。No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Organize AI agents into execution, management, and review roles for complex projects. Choose preset or custom teams, share project context and decisions, and combine their discussions and work into meeting records, feasibility reports, or implementation outputs.
Generate character images, backgrounds, synthetic speech, and music for product videos featuring 小省导购员, a digital shopping presenter. Coordinate these assets with InfiniteTalk to produce a 25-second vertical video made from five scenes.
Convert PDF, Word, and Excel documents into structured Markdown for search and knowledge bases. The skill tries three parsers in sequence when needed and returns metadata, a confidence score, and the parser used.
Organize parsed documents into four color-coded cards covering facts, explanations, risks, and actions. Preserve source references and generation details so readers can review evidence, identify uncertainty, and consider next steps in research, financial, or legal analysis.
Record operations across multiple AI agents and connect events into a traceable history. Search records by tracking ID, time window, or responsible component to support audits and investigations, with local storage available when the vector database fails.
Screen documents and external URLs before they enter a processing workflow. Check blocked domains and document permissions, then produce a pass or reject decision, a risk report, and a tracking ID for later review.