Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/mun1to/vidorq/skillnpx skills add Mun1to/vidorq --skill skillgit clone --depth 1 https://github.com/Mun1to/vidorqWrote 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/mun1to/vidorq/skill)<a href="https://agentmods.dev/skills/mun1to/vidorq/skill"><img src="https://agentmods.dev/badge/skills/mun1to/vidorq/skill.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00000 | $0.01238 |
| Opus 5 | $0.00000 | $0.00619 |
| Sonnet 5 | $0.00000 | $0.00248 |
| Haiku 4.5 | $0.00000 | $0.00124 |
Grade A, and why
skill scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Vidorq skill — edición por IA (v1)
Pipeline de edición que corre desde Claude Code. Hay dos salidas y las dos funcionan: el render directo (PyAV + NVENC), que no necesita Resolve para nada, y el timeline editable dentro de Resolve a través del puente. Auditadas las dos con el mismo encargo el 19-ago-2026. Documentación interna en español (regla D).
Flujo
video crudo ──► transcribe.py ──► transcript.json + takes_packed.md
│
(el LLM lee y razona el corte)
▼
edl.json ──► vidorq_render.py ──► final.mp4
(keep-segments + zoom) (cortes + punch zoom + captions)
Uso
# El intérprete con faster-whisper, PyAV, Pillow y onnxruntime dentro. En una
# instalación normal es el entorno de Vidorq; si usas otro, apúntalo aquí.
PY="$PWD/.venv/Scripts/python.exe"
# 1) Transcribir (word-level, local)
"$PY" skill/helpers/transcribe.py "<video>" "<out_dir>" es
# 2) Autorar edl.json a partir de takes_packed.md (paso de razonamiento del LLM)
# formato: {"strategy": "...", "segments": [{"start","end","zoom","note"}, ...]}
# 3) Renderizar
"$PY" skill/helpers/vidorq_render.py "<video>" "<out_dir>/edl.json" "<out_dir>/transcript.json" "<out_dir>/final.mp4"
# flags: --no-captions --no-zoom
helpers/
- transcribe.py — faster-whisper:
large-v3-turboen float16 sobre la GPU, ysmallint8 en CPU cuando no hay tarjeta o le faltan las librerías de CUDA. Escribetranscript.json(segmentos con timestamps por palabra) ytakes_packed.md(vista compacta para que el LLM razone el corte). - vidorq_render.py — motor de render:
- Cortes: solo los keep-segments del EDL, en orden, con fades de audio de 30 ms en cada frontera (sin pops).
- Punch zoom:
zoompor segmento (p. ej. 1.06) = crop central estático + reescalado. Sin keyframes (respeta la filosofía del MVP). - Captions: chunks Hormozi de 2 palabras UPPERCASE renderizados con PIL (Arial Black, contorno + sombra) y compositados como overlay. Este build de PyAV no trae drawtext/libass.
- Salida vídeo con h264_nvenc (GPU). Vídeo y audio se renderizan por separado y se muxean.
What ships with it
23 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.
- helpers/agente.py 13 KB runs code
- helpers/aprende.py 28 KB runs code
- helpers/autocolor.py 8.3 KB runs code
- helpers/build_resolve_timeline.py 3.2 KB runs code
- helpers/captions.py 51 KB runs code
- helpers/descargar.py 6.4 KB runs code
- helpers/director.py 55 KB runs code
- helpers/efectos.py 12 KB runs code
- helpers/faces.py 8.5 KB runs code
- helpers/fusion.py 15 KB runs code
- helpers/galeria.py 20 KB runs code
- helpers/leer.py 22 KB runs code
- helpers/looks.py 10 KB runs code
- helpers/overlays.py 17 KB runs code
- helpers/previews.py 26 KB runs code
- helpers/providers.py 25 KB runs code
- helpers/speech.py 11 KB runs code
- helpers/transcribe.py 9.3 KB runs code
- helpers/translate.py 10 KB runs code
- helpers/vidorq_render.py 29 KB runs code
- helpers/vision.py 25 KB runs code
- models/face_detection_yunet_2023mar.onnx 227 KB
- models/LICENSE.yunet.txt 1.1 KB
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.
- 4d ago First seen · 78 lines · 0 tokens per session scan A 342e81ebb508
skill is a skill published in the GitHub repository Mun1to/vidorq (5 stars, last pushed 11d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,238 tokens. 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 2026-08-31.
Other skills, from other repositories
canvas
Create, inspect, and update Obsidian JSON Canvas boards with text, file, link, group, and edge nodes. Use for canvas status, canvas lists, visual maps, zones, spatial layouts, adding vault notes or media to a .canvas file, and requests such as create canvas, add to canvas, or put this on the canvas.
youtube-shorts-generator
Generate viral 9:16 YouTube Shorts (or TikTok/Reels clips) from a long-form YouTube URL or local video. Triggers on requests like "make shorts from this video", "extract viral clips from this YouTube link", "auto-clip this podcast", "find the best moments and crop vertical". Pipeline downloads the source, transcribes…
media-processing
Ingest and process media files (video, audio, image).
fish-audio
Generate expressive audio clips using Fish Audio S2 TTS with bracket emotion tags. Record voice memos, narration, audio messages, or any spoken content.
soundclaw
Control Spotify playback, search music, and return shareable music links.
resolve-audio
Audio and Fairlight work in the DaVinci Resolve MCP. Apply when setting audio properties, syncing audio, isolating voice, generating subtitles, planning Fairlight track/bus layouts, checking loudness, routing buses, or splitting/trimming/converting audio — live in a running Resolve OR offline. Routes to the live audio…