[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M77XM4H72CST4F4GXVB78TMC":3,"skill-same:01M77XM4H72CST4F4GXVB78TMC":45},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":20,"frontMatter":21,"refName":22,"commitSha":23,"skillDirectory":24,"forksCount":25,"repositoryHtmlUrl":26,"repositoryHomepageUrl":27,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":29,"updatedAt":30,"install":31,"resources":44},"01M77XM4H72CST4F4GXVB78TMC","video-aroll-auto-editor",null,"Lord1Egypt\u002Fawesome-skill-forge","Lord1Egypt","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F99746359?v=4","01M7592YDNT0XK9WJ0T2MQVPDN","Mohamed Mounir Ali AL Keshita","01M75GRFNHC8VZNWEB8DKDE433",3,"MIT","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fcommunity\u002Fclawhub\u002Fv\u002Fvideo-aroll-auto-editor\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"# Video Auto Editor v4.7\n\nA rule-based and AI-powered automated video roughing tool. It automatically identifies the best segments from original footage and performs editing and splicing.\n\n> **Best for**: **Single-person \u002F talking-to-camera (A'Roll) content** — vlogs, tutorials, podcasts, knowledge-sharing monologues. Not suitable for multi-person dialogues, interviews, or music\u002FB-roll heavy content.\n\n---\n\n## Features\n\n- **Scenario A - Single Video**: Automatically selects the best segment from one video\n- **Scenario B - Batch Processing**: Processes multiple videos, performs cross-video deduplication, and concatenates into one final video\n- **Smart Scoring**: 4-dimension scoring (clear start\u002Fend, fluency, natural rhythm) + fluency analysis\n- **Content Deduplication**: Similarity detection based on transcription text, both within-video and cross-video\n- **Auto Reports**: Generates detailed Markdown reports for each processing run\n\n---\n\n## Best For \u002F Not Suitable\n\n| ✅ Best For | ❌ Not Suitable |\n|-------------|-----------------|\n| Single-person talking to camera (A'Roll) | Multi-person dialogues, interviews |\n| Vlogs, tutorials, podcasts, monologues | Music-heavy, B-roll heavy content |\n| Multiple takes of same content (batch dedup) | Content requiring multiple segments kept |\n| Chinese speech (fluency patterns tuned) | Non-Chinese (patterns not adapted) |\n| Raw long footage (rough cut) | Already tightly edited content |\n\n---\n\n## Requirements\n\n- **Python** 3.8+\n- **FFmpeg** (including ffprobe)\n- **openai-whisper**\n\n### Installation\n\n```bash\n# macOS\nbrew install ffmpeg\npip install openai-whisper\n\n# Ubuntu \u002F Debian\nsudo apt install ffmpeg\npip install openai-whisper\n```\n\n---\n\n## Quick Start\n\n### Scenario A: Single Video\n\nPass a **video file path**:\n\n```bash\npython3 video_editor_auto_v4.6.py .\u002Fvideo.MTS .\u002Foutput\n```\n\nOutput:\n\n```\noutput\u002F\n├── video_粗剪.mp4    # Best segment (clipped)\n└── video_报告.md     # Processing report\n```\n\n### Scenario B: Batch + Deduplication + Concatenation\n\nPass a **folder path** (automatically detected as batch mode):\n\n```bash\npython3 video_editor_auto_v4.6.py .\u002FVideo .\u002Foutput\n```\n\nOutput (only two files, intermediate files are cleaned up):\n\n```\noutput\u002F\n├── 最终拼接_20260311_1905.mp4  # Deduplicated concatenated video\n└── 批量处理报告.md              # Batch report (segment details + dedup decisions)\n```\n\n### Command Format\n\n```\npython3 video_editor_auto_v4.6.py \u003Cinput> [output_dir] [work_dir]\n```\n\n| Parameter | Description | Default |\n|------------|-------------|---------|\n| `\u003Cinput>` | Video file path (Scenario A) or folder path (Scenario B) | Required |\n| `[output_dir]` | Output directory for clips and reports | `.\u002Foutput` |\n| `[work_dir]` | Temporary directory for intermediate files | `.\u002Fvideo_work` |\n\nSupported formats: `.MTS`, `.mp4`, `.mov`\n\n---\n\n## Processing Pipeline\n\n### Scenario A (Single Video)\n\n```\nInput video → Silence detection → Segment identification → 4-dimension scoring\n→ Candidate filtering → Whisper transcription → Fluency analysis\n→ Within-video dedup → Layered selection → Clip output\n```\n\n### Scenario B (Batch)\n\n```\nInput directory → Process each video (Scenario A, no individual reports)\n→ Cross-video deduplication → Concatenate by filename order\n→ Clean intermediate files → Generate single batch report\n```\n\n---\n\n## Configuration\n\nAll parameters are in the `CONFIG` dict at the top of the script:\n\n```python\nCONFIG = {\n    # Silence detection\n    \"silence_noise\": -30,           # dB, lower = stricter\n    \"silence_duration\": 0.8,        # seconds, minimum silence length\n\n    # Filtering\n    \"min_score\": 90,                # Minimum base score (max 100)\n    \"min_duration\": 15,             # Minimum segment duration (seconds)\n\n    # Clip buffer\n    \"buffer_start\": 1,              # Buffer before start (seconds)\n    \"buffer_end\": 3,                # Buffer after end (seconds)\n\n    # Encoding\n    \"crf\": 18,                      # Video quality (18=visually lossless, 23=default)\n    \"preset\": \"fast\",               # Encoding speed\n    \"audio_bitrate\": \"192k\",        # Audio bitrate\n\n    # Adjusted score weights\n    \"penalty_repeat\": 5,            # Per repeat penalty\n    \"penalty_stutter\": 3,           # Per stutter penalty\n    \"penalty_interrupt\": 10,        # Sudden interruption penalty\n    \"bonus_natural_end\": 5,         # Natural ending bonus\n    \"bonus_completeness_max\": 3,    # Completeness bonus cap\n\n    # Deduplication\n    \"duplicate_threshold\": 0.7,     # Content similarity threshold (0-1)\n}\n```\n\n### Tuning Tips\n\n| Scenario | Parameter | Suggested Value |\n|----------|-----------|-----------------|\n| Noisy environment | `silence_noise` | `-35` |\n| Segments too fragmented | `silence_duration` | `1.0` |\n| Want more candidates | `min_score` | `85` |\n| Want shorter segments | `min_duration` | `10` |\n| Higher quality | `crf` | `15` (larger files) |\n\n---\n\n## Scoring System\n\n### Base Score (4 dimensions × 25 points = 100)\n\n| Dimension | Max | Criteria |\n|-----------|-----|----------|\n| Clear start | 25 | Sufficient silence before segment |\n| Clear end | 25 | Sufficient silence after segment |\n| Mid fluency | 25 | Fewer internal interruptions |\n| Natural rhythm | 25 | Low pause ratio + no overly long pauses + not too short |\n\n### Adjusted Score (0-100)\n\nApplied on top of base score based on transcription analysis:\n\n| Item | Points | Description |\n|------|--------|-------------|\n| Repeat penalty | -5 each | \"Re-said\" type stutters (normalized per 30s) |\n| Stutter penalty | -3 each | Filler words (um, uh, etc.) |\n| Interruption penalty | -10 | Ends with connective words (then, but, etc.) |\n| Natural end bonus | +5 | Complete sentence, question, or summary ending |\n| Completeness bonus | +0~3 | Natural end + duration near 60s |\n\n### Layered Selection (Choosing Best Segment)\n\nNot simply the highest score; prioritized filtering:\n\n```\nLayer 1: Prefer naturally ending segments\nLayer 2: Sort by fluency (tolerance 1.5 per 30s)\nLayer 3: Sort by adjusted score\nLayer 4: Tie-break → incomplete: pick last; complete: pick longest\n```\n\n### Deduplication Rules\n\nSame selection rule for within-video and cross-video dedup:\n\n```\nNatural end > Adjusted score > Index\u002Ffilename order (later preferred)\n```\n\n---\n\n## FAQ\n\n### Q: No silence segments detected?\n\nBackground noise may cause misdetection. Try lowering `silence_noise` from `-30` to `-35`.\n\n### Q: Segments cut too finely?\n\nSilence duration threshold may be too short. Increase `silence_duration` from `0.8` to `1.0` or `1.5`.\n\n### Q: Why wasn't the highest-scoring segment chosen?\n\nThe system uses **layered selection**, not raw score comparison. Natural end > Fluency > Adjusted score > Duration. A 95-point segment that ends abruptly may rank lower than a 90-point segment with a natural ending.\n\n### Q: Whisper transcription inaccurate?\n\nDefault uses `small` model for speed\u002Faccuracy balance. To improve:\n- In `transcribe_segment`, change `--model small` to `--model medium` or `--model large`\n- `medium` model has ~85% accuracy for Chinese; recommended if resources allow\n\n### Q: How does cross-video dedup decide which to keep?\n\nSelection rule: Natural end > Adjusted score > Later filename (usually last take, best state).\n\n### Q: Where are the detailed reports?\n\n- Scenario A: `output\u002F\u003Cvideo_name>_报告.md`\n- Scenario B: `output\u002F批量处理报告.md` (includes segment details, transcription summaries, dedup decisions)\n  - Scenario B does not keep intermediate reports or clips; they are cleaned after concatenation\n\n---\n\n## Project Structure\n\n```\nvideo_editor_v4.6_release\u002F\n├── video_editor_auto_v4.6.py   # Main script\n├── README.md                   # This doc\n├── CODE_DOCUMENTATION.md       # Technical doc (architecture, modules, API)\n├── requirements.txt            # Python dependencies\n├── LICENSE                     # GPL v3 license\n└── .gitignore                  # Git ignore rules\n```\n\n---\n\n## Technical Documentation\n\nFor module implementation details, data structures, algorithms, and extension guides, see `CODE_DOCUMENTATION.md`.\n\n---\n\n## License\n\nThis project is licensed under **GPL v3**. If you use this software or its derivatives for commercial purposes, you must release your product as open source under the same license (GPL v3 copyleft requirement).\n\nSee the [LICENSE](LICENSE) file for details.\n\n---\n\n**Version**: v4.7 | **Last Updated**: 2026-03-11\n",{},"master","4aa004da461d820f4ea5499e58eb8d196cf99016","community\u002Fclawhub\u002Fv\u002Fvideo-aroll-auto-editor",0,"https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge","https:\u002F\u002Fskill-explorer-rho.vercel.app","https:\u002F\u002Fgithub.com\u002FLord1Egypt","Unbreakable. Unstoppable. Unmatched⚡.\r\n Speaker of machine tongues 🤖  | World-builder of decentralized dreams 🌐 | Defender of the digital realm🛡️.","2026-06-10 08:21:01",{"source":7,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add Lord1Egypt\u002Fawesome-skill-forge --skill video-aroll-auto-editor",{"tool":37,"command":38},"bunx","bunx skills add Lord1Egypt\u002Fawesome-skill-forge --skill video-aroll-auto-editor",{"tool":40,"command":41},"pnpm","pnpm dlx skills add Lord1Egypt\u002Fawesome-skill-forge --skill video-aroll-auto-editor","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fcommunity\u002Fclawhub\u002Fv\u002Fvideo-aroll-auto-editor\u002FSKILL.md",true,[],[46,53,60,67,74,81],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":18},"01M785WKJB99G3STR1NF9NRTJW","youtube-content","YouTube transcripts to summaries, threads, blogs.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fyoutube-content\u002FSKILL.md",[],[],{"skillNo":54,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":18},"01M785WEFHB0G2S9YA6EPR3H0E","xurl","X\u002FTwitter via xurl CLI: post, search, DM, media, v2 API.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fxurl\u002FSKILL.md",[],[],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":18},"01M785W98EDN4EFJXJPMQEHCRC","touchdesigner-mcp","Control a running TouchDesigner instance via twozero MCP — create operators, set parameters, wire connections, execute Python, build real-time visuals. 36 native tools.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Ftouchdesigner-mcp\u002FSKILL.md",[],[],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":18},"01M785W5S9FP3QBH2V191E025P","test-driven-development","TDD: enforce RED-GREEN-REFACTOR, tests before code.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Ftest-driven-development\u002FSKILL.md",[],[],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":18},"01M785W2AS08NSSH9YATCV6FJB","teams-meeting-pipeline","Operate the Teams meeting summary pipeline via Hermes CLI — summarize meetings, inspect pipeline status, replay jobs, manage Microsoft Graph subscriptions.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fteams-meeting-pipeline\u002FSKILL.md",[],[],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":18},"01M785VXKW4T34BTP1SCHW1R7F","systematic-debugging","4-phase root cause debugging: understand bugs before fixing.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fsystematic-debugging\u002FSKILL.md",[],[]]