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 datit309/supergraph --skill video-demogit clone --depth 1 https://github.com/datit309/supergraphWrote 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/datit309/supergraph/video-demo)<a href="https://agentmods.dev/skills/datit309/supergraph/video-demo"><img src="https://agentmods.dev/badge/skills/datit309/supergraph/video-demo/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/datit309/supergraph/video-demo"><img src="https://agentmods.dev/badge/skills/datit309/supergraph/video-demo.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 | $0.00112 | $0.02331 |
| Opus 5.5 | $0.00045 | $0.00932 |
| Sonnet 5.5 | $0.00022 | $0.00466 |
| Haiku 4.5 | $0.00011 | $0.00233 |
Grade A, and why
video-demo 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 — 156 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/supergraph:video-demo
Films a web app as narrated, zooming, cursor-driven demo videos with retina screenshots, either from a repo it builds and mocks or from a live URL as it is served. Powered by Kokoro local voice-over, smooth camera zooms, natural cursor paths, and Playwright Chromium.
Announce: "🎬 /supergraph:video-demo — filming narrated, zooming, cursor-driven demo videos with retina screenshots..."
Invocations
- Claude Code & Codex:
/supergraph:video-demo [instructions] - DeepSeek Harness (DSH) & Antigravity:
/video-demo [instructions]orskill(name="video-demo") - CLI Binary:
sg-video-demoornpx supergraph sg-video-demo
Overview
scripts/demo films a web app and writes demo-out/ into that project's
workspace: per scene an mp4 with voice-over, a poster, retina stills, plus
full-tour.mp4, manifest.json and index.html. All local. The engine runs
from this skill; nothing is installed into the app being filmed.
Non-negotiables
Each of these cost a run or a user's patience.
- A pinned build, never a dev server. Its module graph belongs to whoever
is editing source. Copy it somewhere private and verify its API base by
reading the bundle. A URL with no repo behind it is
sitemode; a URL on localhost is a dev server and is never filmed. - In
buildmode, mock every request in-process and pin the clock. A corpus gets published, and a real account publishes real people's addresses.sitemode trades both away knowingly: public pages only. - Own your port and output directory. The studio refuses to adopt a server
it did not start. Never
pkillby pattern. - 1080p60 video, 2x stills. They are separate settings. 4K video judders because throughput is a fixed pixels-per-second budget.
- Do the gesture, then narrate while pushed in on the result. One subject per line. Never narrate before the subject exists. Where the app has its own timer — an undo window, a toast — the gesture goes inside it and the words after, or the film shows a feature that does not work.
- One scene, then stop and ask (step 8). It caught four defects before twenty-six more scenes were filmed.
- Read the PNG, every time. A scene that passes while framing the wrong thing is a failure, and only an eye on the still catches it.
- Nothing is final until they say so. Any note about the words, the framing, the pace, what is covered or the voice is a re-film, not a negotiation: change it and film it again. They cannot see what you can, so "too long", "zoom in more" and "stop zooming so much" are all complete instructions, and a scene is cheap to shoot twice.
What ships with it
40 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.
- engine/pin-build.mjs 4.0 KB runs code
- engine/playwright.demo.ts 4.8 KB runs code
- engine/run.mjs 5.1 KB runs code
- engine/serve.mjs 3.7 KB runs code
- engine/studio/actor.ts 23 KB runs code
- engine/studio/config.ts 13 KB runs code
- engine/studio/define.ts 5.9 KB runs code
- engine/studio/overlay.ts 12 KB runs code
- engine/studio/preview.ts 7.0 KB runs code
- engine/studio/ready.ts 3.6 KB runs code
- engine/studio/recorder.ts 23 KB runs code
- engine/studio/run-id.ts 1.8 KB runs code
- engine/studio/scene.ts 6.0 KB runs code
- engine/studio/stage.ts 5.4 KB runs code
- engine/studio/stitch.ts 11 KB runs code
- engine/studio/voice.ts 15 KB runs code
- examples/roadmap-search-poster.png 187 KB
- LICENSE 1.0 KB
- package-lock.json 2.2 KB
- package.json 407 B
- README.md 1.6 KB
- references/capture.md 11 KB
- references/config.md 14 KB
- references/mocking.md 8.8 KB
- references/narration.md 8.1 KB
- references/runs.md 15 KB
- references/scenes.md 10 KB
- references/troubleshooting.md 20 KB
- scripts/demo 303 B
- scripts/doctor.sh 4.6 KB runs code
- scripts/gaps.mjs 5.7 KB runs code
- scripts/install-voice.sh 2.1 KB runs code
- scripts/install.sh 3.9 KB runs code
- scripts/probe-fps.mjs 8.7 KB runs code
- scripts/say.sh 2.3 KB runs code
- scripts/scaffold.mjs 11 KB runs code
- template/demo/demo.config.ts 6.0 KB runs code
- template/demo/mock.ts 4.1 KB runs code
- template/demo/scenes/smoke.scene.ts 1.9 KB runs code
- template/demo/ui.ts 1.1 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.
- 4d ago First seen · 156 lines · 112 tokens per session scan A 2fcce29d9fa1
video-demo is a skill published in the GitHub repository datit309/supergraph (23 stars, last pushed 5d ago), licensed MIT. It adds 112 tokens to every session and 2,331 once invoked, about $0.0004 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 2026-10-24.
Other skills, from other repositories
playwright-cli
Automates browser interactions for testing and validating your own web applications using playwright-cli. Use when you need terminal-first browser control for navigation, form filling, screenshots, tracing, bound browser sessions, debugging, or generating Playwright test code. Only use against applications you own or…
playwright-qa
QA engineering workflows on top of Playwright. Use when a QA engineer asks to design test cases from a ticket, PRD or acceptance criteria, decide what to automate, write a bug report from a failing test or trace, convert manual test cases (TestRail, Xray, CSV) into Playwright specs with case-ID traceability, or review…
safari-extension-control-workflow
Guide macOS Safari integration decisions across Safari Web Extensions, Safari Web Inspector Extensions, Safari App Extensions, SafariServices APIs, content blockers, extension messaging, and app-to-Safari control surfaces. Use when the user wants a Mac app to extend, message, open, inspect, debug, or coordinate with…
safari-mcp-workflow
Inspect a scoped live site in Safari through Safari Technology Preview's local MCP server. Use for Safari-specific DOM, console, network, screenshot, accessibility, performance, interaction, responsive, or print-media evidence.
playwright-skill
Current Playwright (1.64) patterns for writing, debugging, and scaling reliable test suites. Use for any Playwright task, even ones you think you know, because the guides cover APIs and CLI flags newer than most training data. Covers writing or fixing tests, locators, a failing or flaky test, a trace.zip or error…
playwright-migration
Step-by-step migration guides for moving to Playwright from Cypress or Selenium/WebDriver — command mappings, architecture changes, and incremental adoption strategies.