Borrowing it
Nothing to install: this file belongs to parawanderer/OpenTagViewer. 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/parawanderer/OpenTagViewer/main/.claude/skills/device-screenshots/SKILL.mdgit clone --depth 1 https://github.com/parawanderer/OpenTagViewerWrote 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/parawanderer/opentagviewer/device-screenshots)<a href="https://agentmods.dev/skills/parawanderer/opentagviewer/device-screenshots"><img src="https://agentmods.dev/badge/skills/parawanderer/opentagviewer/device-screenshots/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/parawanderer/opentagviewer/device-screenshots"><img src="https://agentmods.dev/badge/skills/parawanderer/opentagviewer/device-screenshots.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00057 | $0.01229 |
| Opus 5 | $0.00028 | $0.00615 |
| Sonnet 5 | $0.00011 | $0.00246 |
| Haiku 4.5 | $0.00006 | $0.00123 |
Grade A, and why
device-screenshots 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 9d 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 — 120 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Looking at the app's UI
Two halves: getting a screenshot off the device, and reading it cheaply. The second is the part that is easy to get wrong.
Never Read a raw screenshot
A device screenshot is ~1080px wide and mostly whitespace. Reading one costs a lot of vision tokens for a picture whose informative part is a few hundred pixels, and reading a before/after pair separately doubles it for a comparison that wants to be side by side anyway.
Run them through the sheet tool first:
python .claude/skills/device-screenshots/sheet.py \
app/build/outputs/managed_device_android_test_additional_output/debug/testEmulator \
<scratchpad>/shots
Then Read the sheets it prints. One subject, all its variants, one image, a few hundred pixels wide.
Files are grouped on the part before the last hyphen; the part after is the variant label:
my_device_list_item-fixed.png -> subject my_device_list_item, variant "fixed"
my_device_list_item-wallpaper.png variant "wallpaper"
| Flag | Default | For |
|---|---|---|
--width |
420 | target width per panel |
--max-height |
420 | cap, so a tall image cannot produce a 2000px sheet |
--dark |
dark,wallpaper_dark,night |
variants composited over black rather than white |
Needs Pillow (python -m pip install pillow).
Screenshots that go somewhere public
A capture destined for the wiki or an issue is a screenshot of a real build on a real device, and some of what is on it belongs to whoever took it. The AMap key dialog is the worked example: it prints the package name and signing fingerprint, because that is exactly what AMap's console asks you to paste — correct behaviour, and not something to publish.
py -3 .claude/skills/device-screenshots/blur.py shot.png --band 0.42 0.55
--band takes fractions of the height, so it survives a change of device resolution; repeat it
for more than one strip. Blur rather than crop — a cropped dialog looks like it has fewer fields
than it does, and the next person wonders what was removed. It pixelates before blurring, so the
characters are gone rather than merely soft.
What ships with it
2 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.
- 9d ago First seen · 120 lines · 57 tokens per session scan A 0f9c68648916
device-screenshots is a skill published in the GitHub repository parawanderer/OpenTagViewer (379 stars, last pushed 2d ago), licensed MIT. It adds 57 tokens to every session and 1,229 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
compose-component-design
Use when designing or reviewing reusable Jetpack Compose component APIs with modifier parameters, root layout placement, caller-provided variable content, primitive content parameters, optional content, or boolean shape flags.
maui-accessibility
Improve MAUI accessibility. USE FOR: semantic labels, hints, headings, screen-reader focus/announcements, AutomationProperties, touch targets, decorative content, TalkBack/VoiceOver/Narrator checks. DO NOT USE FOR: general layout, automation-only tests, performance.
flutter-design-system
Enforce Design Language System adherence in Flutter. Use when implementing design tokens, preventing hardcoded colors/spacing, or building a DLS.
navigate
Call the unintent tool with the following exact parameters.
android-jetpack-compose-expert-v2
Android Jetpack Compose Expert workflow skill. Use this skill when the user needs Expert guidance for building modern Android UIs with Jetpack Compose, covering state management, navigation, performance, and Material Design 3 and the operator should preserve the upstream workflow, copied support files, and provenance…
m3-expressive
Material 3 Expressive design patterns for Jetpack Compose - expressive theming, motion physics, shape morphing, typography emphasis, color emphasis, and all 28 expressive components.