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 Freespirits/social-auto-engine --skill youtube-thumbnailgit clone --depth 1 https://github.com/Freespirits/social-auto-engineWrote 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/freespirits/social-auto-engine/youtube-thumbnail)<a href="https://agentmods.dev/skills/freespirits/social-auto-engine/youtube-thumbnail"><img src="https://agentmods.dev/badge/skills/freespirits/social-auto-engine/youtube-thumbnail/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/freespirits/social-auto-engine/youtube-thumbnail"><img src="https://agentmods.dev/badge/skills/freespirits/social-auto-engine/youtube-thumbnail.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 | $0.00099 | $0.01243 |
| Opus 5 | $0.00049 | $0.00622 |
| Sonnet 5 | $0.00020 | $0.00249 |
| Haiku 4.5 | $0.00010 | $0.00124 |
Grade A, and why
youtube-thumbnail 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.
This is a copy
100% identical to youtube-thumbnail — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 133 lines — stays where its author put it; the contents beside it link to each section on GitHub.
YouTube Thumbnail
CRITICAL: Auto-start on load
When this skill triggers, go straight to Step 1.
Step 1. Gather inputs
Check the project for a reference photo config. Look in this order:
thumbnail-config.mdin the project rootbrand-kit.md— look for a reference image path and brand coloursabout-me.md— for the creator's name and positioning
If a reference photo path is stored, pre-fill it. Otherwise ask:
Upload or provide the path to the reference photo of yourself you want used in the thumbnail. Ideally a clear headshot with distinctive lighting and expression you plan to reuse across videos for brand consistency.
Then call AskUserQuestion:
[
{
"question": "What is the video title?",
"header": "Title",
"multiSelect": false,
"options": [
{"label": "I will type the title", "description": "Type the full working title"},
{"label": "Suggest one", "description": "Given the topic, propose 3 click-worthy titles first"}
]
},
{
"question": "Emotional tone?",
"header": "Tone",
"multiSelect": false,
"options": [
{"label": "Shock / surprise", "description": "Wide eyes, open mouth, bold reaction"},
{"label": "Curious / thinking", "description": "Slight smirk, raised eyebrow, gaze off-frame"},
{"label": "Confident / direct", "description": "Eye contact, calm, assertive"},
{"label": "Frustrated / strong take", "description": "Intense gaze, hand gesture, tension"}
]
}
]
Step 2. Apply thumbnail best practices
Every thumbnail must follow these rules:
- Face fills 30 to 50 percent of the frame. Readable at small sizes.
- 3 to 5 words maximum of large text. 6 if absolutely necessary.
- Two colours dominate. Brand primary + one high-contrast accent (yellow, red, cyan work well).
- One clear focal element besides the face. Tool logo, bold number, arrow, or prop.
- High contrast between face, text, and background. Test by squinting.
- Text is not a sentence. It is a hook phrase. Examples: "I fired my team", "Claude can now...", "Don't do this".
- No small text, no logos bottom-right (watch time icon sits there).
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 · 133 lines · 99 tokens per session scan A 8a14346d43a0
youtube-thumbnail is a skill published in the GitHub repository Freespirits/social-auto-engine (25 stars, last pushed 1mo ago), licensed MIT. It adds 99 tokens to every session and 1,243 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to youtube-thumbnail, differing in 0 lines, and is treated as a copy.
Other skills, from other repositories
ops-ar
OPS on-demand: This skill should be used when the user asks to "A&R this track", "demo verdict", or…
Video Editor (FFmpeg)
AI-powered Video Editor with trimming, merging, effects, overlays, and export.
Video Studio
Remove the original burned-in subtitles via interpolation, analyze channels, and an approval-gated reup pipeline.
Content Video
Turn what the agent read and watched today into a narrated video: script → Content Studio storyboard → images → voice → mp4, as a Codex task.
media_library
A shared bag of raw material — images, GIFs and videos — that every extension can draw from. The user gathers material once; extensions point at a collection by id and the machine picks a file.
aov-mingyu-api
An interface to the aov.cc public API for Chinese astrology, divination, tarot, calendars, and related prompt generation.