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 zhitongblog/unflick --skill unflick-screenshotgit clone --depth 1 https://github.com/zhitongblog/unflickWrote 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/zhitongblog/unflick/unflick-screenshot)<a href="https://agentmods.dev/skills/zhitongblog/unflick/unflick-screenshot"><img src="https://agentmods.dev/badge/skills/zhitongblog/unflick/unflick-screenshot/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/zhitongblog/unflick/unflick-screenshot"><img src="https://agentmods.dev/badge/skills/zhitongblog/unflick/unflick-screenshot.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 | $0.00070 | $0.00559 |
| Opus 5.5 | $0.00028 | $0.00224 |
| Sonnet 5.5 | $0.00014 | $0.00112 |
| Haiku 4.5 | $0.00007 | $0.00056 |
Grade A, and why
unflick-screenshot 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 2mo 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.
What it actually says
unflick: screenshot & picture tuning
Screenshot
screenshot(output?) captures the current frame. With no output a path is
auto-generated; report it back. To screenshot a specific moment, seek() there
first, then screenshot().
unflick seek 42
unflick screenshot --output frame.png
Picture filters
Adjust the live image with the filter_* tools (values −100..100, 0 = neutral):
filter_list()— current brightness/contrast/saturation/gamma/hue.filter_set(name, value)—name∈ {brightness, contrast, saturation, gamma, hue}.filter_reset()— back to neutral.
Looking at the frame yourself
describe_frame(position?, max_edge?) returns the frame as an image block
rather than a path, so a multimodal model can answer "what is on screen right
now" without a round trip through the filesystem. max_edge caps the long side
— keep it modest (768–1024) unless detail is the point.
unflick frame capture --position 42 --max-edge 1024
Geometry
Separate from the colour filters below: video_transform_get() /
video_transform_set(name, value) / video_transform_reset() cover aspect,
rotate, zoom, panscan and deinterlace — the answers to "this is
squashed", "this was filmed sideways" and "this is combing".
Flows
- "Save the current frame" →
screenshot()→ return the path. - "What's happening in this scene?" →
describe_frame()and answer from the image. - "This looks washed out" → bump
filter_set("contrast", +20)and/orfilter_set("saturation", +15); offerfilter_reset()to undo.
Guidance
- Filters affect on-screen playback live; a screenshot taken afterward reflects them. Mention this if the user wants a clean still.
- Keep filter nudges modest (±10–25) and report the resulting values.
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.
- 2mo ago First seen · 63 lines · 70 tokens per session scan A 63c51fe949b4
unflick-screenshot is a skill published in the GitHub repository zhitongblog/unflick (1 stars, last pushed 1mo ago), licensed MIT. It adds 70 tokens to every session and 559 once invoked, about $0.0003 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 2026-09-09.
unflick-screenshot compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| unflick-screenshotzhitongblog/unflick | Skill | 1 | 1mo ago | MIT | A | 70 | Claude Code, Codex |
| imagegenopenai/codex | Skill | 134k | today | Apache-2.0 | A | 113 | Codex |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 79k | yesterday | MIT | A | 200 | Claude Code, Codex |
| playwright-climicrosoft/playwright | Skill | 100k | yesterday | Apache-2.0 | A | 19 | Claude Code |
| browser-automationopenclaw/openclaw | Skill | 394k | today | MIT | A | 36 | Claude Code |
| systematic-debuggingobra/superpowers | Skill | 314k | 4d ago | MIT | A | 21 | Claude Code |
| next-cache-components-adoptionvercel/next.js | Skill | 143k | today | MIT | A | 95 | Claude Code |