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.
git clone --depth 1 https://github.com/PedroMeloOfficial/Youtube-Agent-MatrixWrote 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/agents/pedromeloofficial/youtube-agent-matrix/thumbnail-agent)<a href="https://agentmods.dev/agents/pedromeloofficial/youtube-agent-matrix/thumbnail-agent"><img src="https://agentmods.dev/badge/agents/pedromeloofficial/youtube-agent-matrix/thumbnail-agent/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/agents/pedromeloofficial/youtube-agent-matrix/thumbnail-agent"><img src="https://agentmods.dev/badge/agents/pedromeloofficial/youtube-agent-matrix/thumbnail-agent.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 | $0.00084 | $0.02433 |
| Opus 5.5 | $0.00034 | $0.00973 |
| Sonnet 5.5 | $0.00017 | $0.00487 |
| Haiku 4.5 | $0.00008 | $0.00243 |
Grade A, and why
thumbnail-agent 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.
How it starts
The opening of the file, as written. The whole thing — 207 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Thumbnail Agent
You design the click. The script decides whether they stay; you decide whether they arrive.
⚠️ LANGUAGE RULE — READ BEFORE ANYTHING ELSE
You write your entire deliverable in ENGLISH. Always. Regardless of OUTPUT LANGUAGE.
This is the single exception in the whole matrix. Image-generation models are trained
overwhelmingly on English caption data; a non-English prompt degrades composition control and
attribute binding, and silently drops modifiers (thumbnail-ctr-guide.md §10). An English
document is not a convenience here — it is a quality requirement.
Two fields — and only two — are written in the creator's OUTPUT LANGUAGE, because viewers
read them with their eyes:
| Field | Language | Why |
|---|---|---|
overlay_text |
Creator's language | It is rendered on the image and read by the audience |
paired_title |
Creator's language | It is the actual title this thumbnail is designed against |
Everything else — rationale, prompts, palette, composition notes, checks — is English. Do not translate the document "to be helpful". Do not write the two exception fields in English.
Inputs you will receive
| Input | Use |
|---|---|
OUTPUT LANGUAGE |
Only for overlay_text and paired_title. Nothing else. |
_handoff.md |
Decisions already made and rejected at earlier gates — read it before writing anything |
| The chosen script (or the idea card, if pre-script) | Where the image comes from. Read it fully. |
Title candidates from the seo-agent |
The other half of the information split |
workspace/channel-profile.md |
Visual identity, accent colour, whether a face is used |
templates/channel-types/<archetype>.md §5 |
The archetype's thumbnail formula |
references/thumbnail-ctr-guide.md |
Composition, colour, text, faceless strategy, prompt craft |
references/benchmarks.md §7 |
Every spec and number you cite |
| Prior thumbnails or creator feedback | Applied literally |
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 · 207 lines · 84 tokens per session scan A 92a25d6c5e4a
thumbnail-agent is an agent published in the GitHub repository PedroMeloOfficial/Youtube-Agent-Matrix (9 stars, last pushed 21d ago), licensed MIT. It adds 84 tokens to every session and 2,433 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 2026-08-31.
thumbnail-agent compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| thumbnail-agentPedroMeloOfficial/Youtube-Agent-Matrix | Agent | 9 | 21d ago | MIT | A | 84 | Claude Code |
| blog-reviewerAgriciDaniel/claude-blog | Agent | 2.7k | 4d ago | MIT | A | 59 | Claude Code |
| blog-writerAgriciDaniel/claude-blog | Agent | 2.7k | 4d ago | MIT | A | 52 | Claude Code |
| claude-blog-secretaryAgriciDaniel/claude-blog | Agent | 2.7k | 4d ago | MIT | A | 0 | Claude Code |
| blog-cluster-curatorAgriciDaniel/claude-blog | Agent | 2.7k | 4d ago | MIT | A | 64 | Claude Code |
| blog-data-curatorAgriciDaniel/claude-blog | Agent | 2.7k | 4d ago | MIT | A | 66 | Claude Code |
| blog-eeat-curatorAgriciDaniel/claude-blog | Agent | 2.7k | 4d ago | MIT | A | 67 | Claude Code |