[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C27VCGGJVJQJS66S1E4QD5":3,"skill-same:01M7C27VCGGJVJQJS66S1E4QD5":49},{"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":6,"ownerHtmlUrl":27,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":28,"install":29,"resources":42},"01M7C27VCGGJVJQJS66S1E4QD5","silence-detector",null,"linyouxue\u002Fskill-repair-benchmark","linyouxue","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F216151464?v=4","01M7BJEVCXKD74KH6FG9CBVSQB","林宥学","01M7BJEVE02NP217V93BKZFC0J",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FSkillrevise\u002Ftasks\u002Fvideo-silence-remover\u002Frepaired_skill\u002Fsilence-detector\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"# Pause-and-Silence Remover (Media)\n\n## Purpose\nDetect long pauses\u002Fsilence throughout an audio\u002Fvideo file, remove those intervals, and produce an edited media file plus a machine-readable report, with verifier-aligned output checks.\n\n## When to Use\nUse when a task requires finding pauses\u002Fsilent gaps across the full timeline (not just the beginning), removing intervals longer than a specified minimum duration, and emitting explicit output artifacts (edited media and a JSON report) at paths named in the task instructions.\n\n## Procedure\n- Step 1: Preflight (inputs, outputs, tools, assumptions)\n- Read the task instructions and extract: input media path, required output media path, required report path, and the minimum pause duration threshold (if unspecified, stop and request clarification rather than guessing).\n- Validate input media path: exists, is a regular file, readable, and non-zero size.\n- Discover tools (do not assume fixed install locations):\n- Prefer `ffmpeg` and `ffprobe` if available (`command -v ffmpeg`, `command -v ffprobe`).\n- If missing, stop and request an alternative tool allowance rather than inventing commands.\n- Execution anchor: run `ffprobe` to record baseline duration and confirm decodability; keep the measured duration for later bounds checks.\n- Step 2: Detect candidate pause segments across the entire timeline (choose a method, then normalize)\n- Primary method (preferred when available): ffmpeg `silencedetect` on the audio track to find silence start\u002Fend events.\n- Run detection to produce a plain-text log (to a temp file in the working directory) and parse it into segments.\n- Fallback method: if `silencedetect` is unavailable or produces no usable events, use an energy-based approach (frame\u002Fwindow energy with smoothing) only if the needed dependencies are present; otherwise stop and request guidance.\n- Normalize segments:\n- Convert detection events into `[start, end]` intervals in seconds.\n- Clamp to `[0, media_duration]`.\n- Drop intervals with `end \u003C= start`.\n- Apply the task-defined minimum duration threshold (keep only intervals with duration >= threshold).\n- Optionally merge intervals separated by a very small gap only if the task specifies merge behavior; otherwise do not merge.\n- Checkpoint: if zero segments remain, do not silently fail; proceed with a no-op edit plan (copy\u002Fremux original) but still generate a valid report stating that no pauses above threshold were found.\n- Step 3: Build a safe cut plan and produce the edited media\n- Construct the keep-intervals as the complement of the removed segments over `[0, duration]`.\n- Validate the plan before editing:\n- Keep-intervals must be sorted, non-overlapping, within bounds, and have positive durations.\n- Total kept duration should be `original_duration - total_removed_duration` within a small tolerance (to account for timestamp rounding).\n- Render edited output using a method appropriate to the environment and requirements:\n- Prefer a concat-based workflow that trims and concatenates keep-intervals in timeline order.\n- Choose stream copy vs re-encode based on task constraints; if not specified, prefer a safe default that produces a broadly playable file (but do not claim lossless behavior).\n- Write output to the exact required output media path from the task (do not redirect to a different directory if it fails).\n- Step 4: Write the report JSON and verify outputs (must pass observable checks)\n- Create a report JSON that (at minimum) includes:\n- `method` (string), `parameters` (object including min_pause_duration and detection settings used),\n- `removed_segments` (list of objects with numeric `start`, `end`, `duration`),\n- `totals` (object with total_removed_duration_seconds, original_duration_seconds, output_duration_seconds).\n- If the task specifies an exact schema\u002Ffield names, follow that schema instead of this minimal one.\n- Post-write validation (schema + bounds):\n- Reload the JSON from disk and validate required keys\u002Ftypes.\n- Assert every segment is within `[0, original_duration]` and `duration == end - start` within tolerance.\n- Execution anchor: final verifier-aligned checks at the exact task-specified output paths:\n- Output media file exists, is readable, and size > 0.\n- `ffprobe` can decode the output media (non-error exit).\n- Report file exists, is readable, and parses as valid JSON.\n- If any check fails, do not switch paths silently; instead diagnose: permissions, missing mounts, wrong output path extracted from instructions, or ffmpeg failure logs.\n\n## Constraints \u002F Pitfalls\n- Never hard-code absolute script paths, repo-internal locations, or assume auxiliary skills exist; always discover available tools and operate from task-provided input\u002Foutput paths.\n- Do not remove or cut media until (a) input decodes via ffprobe, (b) pause threshold is confirmed from instructions, and (c) segments are normalized and validated to be in-bounds and non-overlapping; otherwise you risk corrupt outputs or removing content incorrectly.",{},"main","cdcb6d230f3b15c3c33f8a2bdd16439f43602cbe","evaluation\u002Fresults\u002FSkillrevise\u002Ftasks\u002Fvideo-silence-remover\u002Frepaired_skill\u002Fsilence-detector",1,"https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark","https:\u002F\u002Fgithub.com\u002Flinyouxue","2026-11-14 20:28:35",{"source":7,"skillName":5,"commands":30,"prompt":40,"zipAvailable":41},[31,34,37],{"tool":32,"command":33},"npx","npx skills add linyouxue\u002Fskill-repair-benchmark --skill silence-detector",{"tool":35,"command":36},"bunx","bunx skills add linyouxue\u002Fskill-repair-benchmark --skill silence-detector",{"tool":38,"command":39},"pnpm","pnpm dlx skills add linyouxue\u002Fskill-repair-benchmark --skill silence-detector","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FSkillrevise\u002Ftasks\u002Fvideo-silence-remover\u002Frepaired_skill\u002Fsilence-detector\u002FSKILL.md",true,[43],{"path":44,"resourceType":45,"resourceTypeDesc":46,"sizeBytes":47,"sourceHtmlUrl":48},"scripts\u002Fdetect_silence.py","SCRIPT","脚本",3248,"https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FSkillrevise\u002Ftasks\u002Fvideo-silence-remover\u002Frepaired_skill\u002Fsilence-detector\u002Fscripts\u002Fdetect_silence.py",[50,57,64,71,76,81],{"skillNo":51,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":54,"categories":55,"keywords":56,"publishedAt":18},"01M7E1HGVRY73NBJ022CJG13X6","video-frame-extraction","Extract frames from video files and save them as images using OpenCV","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_03\u002Finputs\u002Fskills\u002Fvideo-frame-extraction\u002FSKILL.md",[],[],{"skillNo":58,"name":59,"nameZh":6,"description":60,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":61,"categories":62,"keywords":63,"publishedAt":18},"01M7E1HAG5R7CJ9C8NYJS771MB","openai-vision","Analyze images and multi-frame sequences using OpenAI GPT vision models","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_03\u002Finputs\u002Fskills\u002Fopenai-vision\u002FSKILL.md",[],[],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":18},"01M7E1H7DADKB6QBWKC2KQ63SZ","image-ocr","Extract text content from images using Tesseract OCR via Python, including structured field extraction from receipts and forms","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_03\u002Finputs\u002Fskills\u002Fimage-ocr\u002FSKILL.md",[],[],{"skillNo":72,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"publishedAt":18},"01M7E1H0WBNT43TENGQY7AVZHD","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_02\u002Finputs\u002Fskills\u002Fvideo-frame-extraction\u002FSKILL.md",[],[],{"skillNo":77,"name":59,"nameZh":6,"description":60,"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},"01M7E1GTC7D4N4Q2EF03875J9E","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_02\u002Finputs\u002Fskills\u002Fopenai-vision\u002FSKILL.md",[],[],{"skillNo":82,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":18},"01M7E1GQ92CWFMGH53KSSE4WER","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_02\u002Finputs\u002Fskills\u002Fimage-ocr\u002FSKILL.md",[],[]]