Borrowing it
Nothing to install: this file belongs to daniel-ddtech/diorama. 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/daniel-ddtech/diorama/main/.claude/skills/record-extension-demo/SKILL.mdgit clone --depth 1 https://github.com/daniel-ddtech/dioramaWrote 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/daniel-ddtech/diorama/record-extension-demo)<a href="https://agentmods.dev/skills/daniel-ddtech/diorama/record-extension-demo"><img src="https://agentmods.dev/badge/skills/daniel-ddtech/diorama/record-extension-demo.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 5.1 | $0.00047 | $0.00888 |
| Opus 5 | $0.00023 | $0.00444 |
| Sonnet 5 | $0.00009 | $0.00178 |
| Haiku 4.5 | $0.00005 | $0.00089 |
Grade A, and why
record-extension-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 7d 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 — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Record an extension demo with Diorama
Diorama records real Chrome extensions headlessly from a YAML beat sheet and composites a finished mp4 with a drawn browser frame, synthetic cursor, click ripples, and camera moves. No screen recording, no OS permissions.
Workflow
- Check the environment:
diorama doctor(ornode <diorama>/packages/cli/dist/main.js doctor). Needs Chrome for Testing (DIORAMA_CHROME or a Playwright Chromium) and ffmpeg. - Find the extension's unpacked directory (contains manifest.json, must be Manifest v3). Read the manifest to learn the popup path and the extension name.
- Write a beat sheet. Start from
diorama initor this shape:
version: 1
title: My extension demo
viewport: { width: 1280, height: 860, scale: 2 } # +100px toolbar = 4:3
extension:
path: /abs/path/to/extension
popup: { width: 600, height: 600, autoSize: true }
steps:
- verb: goto
url: https://the-page-the-extension-works-on.example
- verb: wait
selector: "css-or-text=selector"
- verb: openPopup
- verb: camera
zoom: 1.35
focus: popup
ms: 700
- verb: click
target: popup
selector: text=The Button Label
- verb: wait
target: popup
expression: /done|complete/i.test(document.body.innerText)
- verb: hold
ms: 1200
- verb: mark
name: end
Verbs: goto, wait (selector | ms | expression), click, type, scroll, hover,
openPopup, closePopup, camera (zoom 1-2.5, focus page|popup, ms), hold, mark.
text= selectors match visible text case-insensitively. The last mark is the
poster frame. Optional blocks: frame (theme dark|light|custom html path, url,
title), profile (dir, seedStorage), cursor (scale, ripple, shadow), output
(fps, endCard true|false|{title,subtitle}, posterAt, formats list).
- Record:
diorama record demo.beats.yaml --out out/ --keep-run. - Inspect the result before declaring success: extract frames with ffmpeg
(
ffmpeg -ss <t> -i out/<title>.mp4 -frames:v 1 check.jpg) and view them. Common fixes: add wait steps before clicks that miss, add holds so states breathe, adjust camera timing to land after UI settles. - Iterate by editing the sheet and re-running. Recordings are reproducible.
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.
- 7d ago First seen · 82 lines · 47 tokens per session scan A 18415da46b87
record-extension-demo is a skill published in the GitHub repository daniel-ddtech/diorama (0 stars, last pushed 21d ago), licensed MIT. It adds 47 tokens to every session and 888 once invoked, about $0.0002 per session on Opus 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 2026-08-31.
Other skills, from other repositories
demo
Record a short demo VIDEO of a PR or feature — drive the repo's web UI in a real headless browser, capture the session, and composite a titled, size-capped mp4 with ffmpeg. Use on the docker QA image when the deliverable is a playable demo clip (single walkthrough or before/after comparison), not a text/screenshot…
demo-video
Produce a polished, ScreenStudio-style 60fps product demo video of a web app with a tiny, dependency-light pipeline — Playwright frame capture + a multiprocess Pillow compositor streaming into ffmpeg (no Node/Remotion). Drives the app one step per frame so playback is true 60fps regardless of capture speed, then…
demo-video
A process for automatically recording feature-demo videos of a local web app using Playwright, a browser-automation tool, and ffmpeg, a video-processing tool. It drives the real app, then adds subtitles, labels, a simulated cursor, zooms, and faster playback for waiting periods.
annotation-feedback-wait
A workflow for collecting and waiting for visual feedback that users draw on a web page with a browser annotation tool.
annotation-browser-launch
A workflow for opening an agent-controlled browser with a page-annotation extension so a user can mark up a web page.
cleya-edit-craft
The grammar of cuts — which cut to use, when to cut, and how to pace an ad. Read before authoring any multi-shot film.