Borrowing it
Nothing to install: this file belongs to Enriquemarq1/zernio-library-skills. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Enriquemarq1/zernio-library-skills/main/.claude/skills/clip-machine/SKILL.mdgit clone --depth 1 https://github.com/Enriquemarq1/zernio-library-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/enriquemarq1/zernio-library-skills/clip-machine)<a href="https://agentmods.dev/skills/enriquemarq1/zernio-library-skills/clip-machine"><img src="https://agentmods.dev/badge/skills/enriquemarq1/zernio-library-skills/clip-machine/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/enriquemarq1/zernio-library-skills/clip-machine"><img src="https://agentmods.dev/badge/skills/enriquemarq1/zernio-library-skills/clip-machine.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Data Exfiltration · line 78 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00121 | $0.01456 |
| Opus 5.5 | $0.00048 | $0.00582 |
| Sonnet 5.5 | $0.00024 | $0.00291 |
| Haiku 4.5 | $0.00012 | $0.00146 |
Grade A, and why
clip-machine scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 112 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Clip Machine — one long video in, native posts out
Four stations. Claude runs the first three, Zernio ships the fourth. No timeline software ever opens.
LONG VIDEO → [1 CUT] → [2 CAPTION] → [3 COVER] → [4 POST] → native posts live
The math this kills: 1 video → 3 clips × 3 platforms = 9 manual uploads, every week, forever. With the machine it's 3 requests.
Starting from zero? Copy-paste reference/initiation-prompt.md to Claude
Code in an empty folder — it sets up the whole system for you.
STATION 1 — CUT
- Transcribe with word timestamps. WhisperX is the reference model —
open source, free, runs on your machine (
pip install whisperx, needs ffmpeg):whisperx video.mp4 --model large-v3 --output_format jsonThe JSON gives every word a start/end timestamp via forced alignment (GPU: add--device cuda; CPU works with--compute_type int8). Any word-level transcription service works too — the machine only needs words with millisecond timing. No ground truth, no clean cut. - Pick the best moments with the selection framework in
reference/best-moments-prompt.md. That file IS the prompt — paste it with the transcript. Core law: finish the point — a clip starts at the beginning of a thought and ends after the payoff, never mid-sentence. - Cut on word boundaries. Start ~0.15s before the first word, end
~0.25s after the last (never past the next word's start). ffmpeg:
ffmpeg -ss {start} -to {end} -i long.mp4 -c:v libx264 -crf 18 -c:a aac clip.mp4 - Reframe to 9:16 (1080×1920). Crop to the speaker (face-tracking via MediaPipe if available; a fixed crop works when the framing is static). Screen-share sources: stack the screen crop on top, the face cam below.
STATION 2 — CAPTION (there are TWO captions — that's the teach)
- Burned-in word captions on the video. Most people watch muted. Word-synced from the transcript timestamps (that's why Station 1 needs millisecond words), 3-4 words per line, high-contrast, inside the caption-safe band (keep out of the bottom ~20% and top ~12% — platform UI lives there).
- Caption TEXT per platform — never the same text three times.
Copy-pasting one caption everywhere is the amateur tell. Claude writes
all three from the transcript:
- TikTok: conversational hook + 3-5 niche hashtags.
- Reels: the first ~125 chars must stop the scroll (that's all that shows before "...more").
- Shorts: barely displays text — put the value in the TITLE instead.
Full rules per platform:
reference/platform-cheat-sheet.md.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 112 lines · 121 tokens per session scan A f87585727e08
clip-machine is a skill published in the GitHub repository Enriquemarq1/zernio-library-skills (49 stars, last pushed 3d ago), licensed MIT. It adds 121 tokens to every session and 1,456 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
wechatsync
A publishing tool for sending Markdown or HTML articles to many writing and social platforms, including Chinese services, WordPress, Typecho, and WeChat. Cross-posting means publishing the same article in several places.
kotlin-expert
Expert-level Kotlin development, Android, coroutines, and multiplatform. Use when the user mentions Android, coroutines, multiplatform, or JVM, or when the task involves Kotlin Fundamentals, Android Development, Kotlin Multiplatform, or Kotlin Style.
content-matrix
Generate 32+ LinkedIn post ideas in a single table by pairing the user's content pillars with 8 proven content formats. Based on the Justin Welsh content matrix. Use this skill whenever the user says "give me post ideas", "content matrix", "what should I post about", "generate post ideas", "content ideation", or "map…
instagram-api
An Instagram API alternative on fetcher.sh — pay-per-call in USDC via x402, or prepaid credits with a Bearer key, no login and no session cookies. Use when the user wants to resolve an Instagram profile by @handle, search users by keyword, pull a profile's posts, reels, stories, tagged posts, followers, or followings…
kotlin-docs
Comprehensive Kotlin 2.4.0 reference covering all language features: variables, basic types, strings, control flow, functions, lambdas, classes, objects, inheritance, interfaces, data classes, sealed classes, generics, collections, sequences, null safety, coroutines (suspend, launch, async, Flow, StateFlow, channels)…
socialblast-pipeline
Plan and create a week of social media content for any business using SocialBlast AI's MCP tools. Use this skill when the user says "create a campaign", "generate a week of content", "make me posts for my business", "plan content for", or asks Claude to enrich pending posts with AI media. Drives the SocialBlast MCP…