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 skills add maddexritter-rgb/vibe-editing --skill watchgit clone --depth 1 https://github.com/maddexritter-rgb/vibe-editingWrote 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/maddexritter-rgb/vibe-editing/watch)<a href="https://agentmods.dev/skills/maddexritter-rgb/vibe-editing/watch"><img src="https://agentmods.dev/badge/skills/maddexritter-rgb/vibe-editing/watch/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/maddexritter-rgb/vibe-editing/watch"><img src="https://agentmods.dev/badge/skills/maddexritter-rgb/vibe-editing/watch.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00196 | $0.01816 |
| Opus 5 | $0.00098 | $0.00908 |
| Sonnet 5 | $0.00039 | $0.00363 |
| Haiku 4.5 | $0.00020 | $0.00182 |
Grade B, and why
watch scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads MCP configurationmediumAgent snooping
mcp.json carries server URLs and auth tokens; reading it lets a mod discover and abuse other integrations.
(the project `.mcp.json` reads `${GEMINI_API_KEY}`, or hard-code it there like the How it starts
The opening of the file, as written. The whole thing — 130 lines — stays where its author put it; the contents beside it link to each section on GitHub.
watch — eyes & ears for video
The thing that makes "watch a video well" work is converting video into things Claude reasons over natively: frames as images (Claude sees them directly) + an accurate transcript + timing/loudness metadata. This skill is the disciplined way to do that, tuned for short-form editing QA and reference-matching.
How Claude actually sees video
- Frames → my own eyes. The MCP
video_watchwithframe_mode:"images"(now the default — see config below) returns real JPEGs into the conversation. I look at them directly; there is no lossy "describer" model in between. The local scripts here do the same via ffmpeg → PNG that I then Read. - Audio → transcript. The MCP transcribes with the configured backend. For editing captions, keep using the canonical transcription stack (Groq + WhisperX + vocab-correct) — it is better on proper nouns. Use the MCP transcript for understanding/QA, not for generating final captions.
- Long-form (>~3 min) → Gemini. When the Gemini key is set, flip the backend to
gemini-apiand Gemini ingests the whole video in one pass (timestamps, what happens, where the energy is). Without it, dense local frames + transcript still work, just at higher token cost — so segment first.
Decision tree — which tool, when
| Goal | Use |
|---|---|
| Understand a whole long-form fast | MCP video_analyze (structure) → video_watch Gemini one-pass |
| Eyeball a finished short's arc | contact_sheet.py → Read the PNG |
| Check burnt-in caption placement/presence | caption_ocr.py (placement) + contact sheet & your eyes (exact text) |
| Compare my render vs a reference reel | ab_diff.py --mode side |
| See exactly what changed v1 → v2 | ab_diff.py --mode diff |
| Check audio levels / pacing / black frames | probe.py (or MCP video_analyze loudness/silence) |
| Zoom into one tense moment frame-by-frame | MCP video_detail with a tight segment + high fps |
Default rhythm for a clip I have not seen: contact_sheet.py first (one image, whole
arc) → decide which moments matter → video_detail/video_watch on just those segments at
higher fps. Never blast a 13-min file at high fps; analyze, then drill.
What ships with it
5 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.
- 11d ago First seen · 130 lines · 196 tokens per session scan B 3e8c324d418e
watch is a skill published in the GitHub repository maddexritter-rgb/vibe-editing (7 stars, last pushed 2mo ago), licensed MIT. It adds 196 tokens to every session and 1,816 once invoked, about $0.0010 per session on Opus 5. A static security scan graded it B with 1 finding (reads mcp configuration). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
general-video
Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated…
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
chengfeng-check-updates
An environment manager for a video-editing system. It checks whether its skills and runtime—the software needed to run them—are installed and compatible.
diagnostic-stem-delivery
Audio production with diagnostic analysis, timecode parsing from documents, and verified export workflow.