Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/hub-ki/video-generationnpx agentmods add skills/hub-ki/video-generation/demo-videoWrote 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/hub-ki/video-generation/demo-video)<a href="https://agentmods.dev/skills/hub-ki/video-generation/demo-video"><img src="https://agentmods.dev/badge/skills/hub-ki/video-generation/demo-video/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/hub-ki/video-generation/demo-video"><img src="https://agentmods.dev/badge/skills/hub-ki/video-generation/demo-video.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 | $0.00236 | $0.13189 |
| Opus 5.5 | $0.00094 | $0.05276 |
| Sonnet 5.5 | $0.00047 | $0.02638 |
| Haiku 4.5 | $0.00024 | $0.01319 |
Grade B, and why
demo-video 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 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.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
`ffmpeg-static`/`ffprobe-static` packages in `~/.hyperframes-cli` (all platforms, no sudo — **the How it starts
The opening of the file, as written. The whole thing — 805 lines — stays where its author put it; the contents beside it link to each section on GitHub.
demo-video
You need a finished, designed video of an app. This skill is the tested, batteries-included pipeline. It builds on HyperFrames (HTML→video) but adds the parts that are painful to rediscover: how to record the app yourself, this machine's ffmpeg workarounds, the exact edit recipes, the branded design system, and the traps that waste hours.
Nothing here is branded. The look is one design with swappable tokens; whose brand a video is in is a step, not an assumption — see Phase 0B.
🚦 The phases are grouped by topic, not by order of work
Read this before following anything below in sequence. The capture mechanics are documented first because they belong together, but capture is not the first thing you do:
The concept comes first: what is this FOR? → the arc → PLAN.md (+ SCRIPT.md) → APPROVED
(that is the `video-plan` skill, not this one) │
│
↓ only now does footage exist ──────────────────────────────────────────────────────┘
Phase 0 environment → 0B brand → 1A capture / 1B ingest → 3 prep → 4 compose
→ 5 verify → 6 render
Shooting first looks faster and is not: the arc is what changes under review, and every beat captured against an unagreed arc is re-captured. Phase 0 and 0B are the two things you may run early — they install tools and settle the palette. Everything in Phase 1 waits for an approved plan.
🚀 Trying it out — the fast path
🚨 The caller chooses this mode. You never do. You may OFFER it, in one question. You may not conclude it from circumstances, and these in particular are not consent:
- they attached or unpacked the skill package — that is how someone installs a tool, not a statement about what they want made with it;
- the request sounds exploratory, or came without a brief;
- it is the first thing they have asked for;
- you would find the fast path more convenient.
What ships with it
60 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.
- assets/brand.example.json 430 B
- assets/dpr-fixture.html 1.4 KB
- assets/fixtures/app-iframe-inner.html 493 B
- assets/fixtures/app-iframe.html 445 B
- assets/fixtures/app-shell.html 415 B
- assets/fixtures/banner-on-bare-page.html 1.2 KB
- assets/fixtures/bottom-banner.html 848 B
- assets/fixtures/compact-card.html 794 B
- assets/fixtures/duplicates.html 1004 B
- assets/fixtures/iframe-wall-inner.html 373 B
- assets/fixtures/iframe-wall.html 895 B
- assets/fixtures/in-page-agree.html 694 B
- assets/fixtures/inflow-then-banner.html 1.0 KB
- assets/fixtures/invitation-modal.html 1.1 KB
- assets/fixtures/late-consent.html 1.2 KB
- assets/fixtures/lazy-iframe.html 855 B
- assets/fixtures/long-wall-short-page.html 1.3 KB
- assets/fixtures/modal-variants.html 981 B
- assets/fixtures/native-dialog.html 658 B
- assets/fixtures/plain-agree.html 448 B
- assets/fixtures/responsive-embed-inner.html 344 B
- assets/fixtures/responsive-embed.html 594 B
- assets/fixtures/reveal.html 719 B
- assets/fixtures/scrollable-wall.html 768 B
- assets/fixtures/shell-with-agree.html 762 B
- assets/fixtures/sticky-header-invite.html 904 B
- assets/fixtures/translucent-shell.html 658 B
- assets/fixtures/untranslated-banner.html 627 B
- assets/template.html 12 KB
- examples/starter/narration.mjs 829 B runs code
- examples/starter/PLAN.md 895 B
- examples/starter/README.md 507 B
- examples/starter/SCRIPT.md 794 B
- README.md 12 KB
- references/brand-style.md 14 KB
- references/companion-docs.md 3.6 KB
- references/container-capture.md 6.8 KB
- references/design-system.md 32 KB
- references/ffmpeg-recipes.md 6.5 KB
- references/foreign-sites.md 27 KB
- references/guide-track-build.md 7.0 KB
- references/multiple-takes.md 2.0 KB
- references/pitfalls.md 27 KB
- references/playwright-capture.md 27 KB
- references/timeline-and-review.md 5.0 KB
- references/voiceover.md 12 KB
- scripts/app-state.mjs 12 KB runs code
- scripts/apply-brand.mjs 8.4 KB runs code
- scripts/audit-composition.mjs 24 KB runs code
- scripts/brand-lib.mjs 8.6 KB runs code
- scripts/capture-auth.js 5.5 KB runs code
- scripts/capture-lib.js 86 KB runs code
- scripts/capture-template.js 21 KB runs code
- scripts/crop-spots.mjs 5.3 KB runs code
- scripts/extract-brand.mjs 13 KB runs code
- scripts/make-audio.mjs 6.0 KB runs code
- scripts/make-brand.mjs 9.0 KB runs code
- scripts/measure-speech.mjs 3.4 KB runs code
- scripts/measure-spots.mjs 6.8 KB runs code
- scripts/mux-audio.mjs 3.5 KB runs code
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 · 805 lines · 236 tokens per session scan B 0afbbfa45794
demo-video is a skill published in the GitHub repository hub-ki/video-generation (6 stars, last pushed 2mo ago), licensed MIT. It adds 236 tokens to every session and 13,189 once invoked, about $0.0009 per session on Opus 5.5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
keep-the-why
In a project with a .keep-the-why, or when asked about Keep the Why or to set it up - extract and preserve the reasoning code cannot explain - decisions, rejected alternatives, workarounds, incidents, constraints - plus project setup and maintainer interviews. Not for what changed (see Keep a Changelog) - only why.…
sw-do
Work a SpecWeave increment task by task - claim, implement inside its Files, test, commit, done with evidence - alone, unattended or in lanes, then close. Use for "implement", "next task".
done
Close a SpecWeave increment - every task done or skipped, specweave verify green, a fresh-context review, then specweave complete. Use for "close it", "we are done", "finish up".
review
Adversarial review of an increment before it ships, by a fresh context or another model family; every finding cites path:line and is re-verified. Use for "review", "qa", "critique this".
nextjs-supabase-vercel
Set up the current empty folder with the AI App Starter, so the agent builds a Next.js, Supabase and Vercel app with the owner step by step, from idea to launch.
summarize
A guide for turning conversations, logs, documents, or investigation notes into short, action-oriented summaries. It requires conclusions to be tied to available evidence and distinguishes confirmed facts from items still needing confirmation.