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 chanhoan/anime-persona-skills --skill maringit clone --depth 1 https://github.com/chanhoan/anime-persona-skillsWrote 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/chanhoan/anime-persona-skills/marin)<a href="https://agentmods.dev/skills/chanhoan/anime-persona-skills/marin"><img src="https://agentmods.dev/badge/skills/chanhoan/anime-persona-skills/marin/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/chanhoan/anime-persona-skills/marin"><img src="https://agentmods.dev/badge/skills/chanhoan/anime-persona-skills/marin.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 | $0.00105 | $0.01821 |
| Opus 5.5 | $0.00042 | $0.00728 |
| Sonnet 5.5 | $0.00021 | $0.00364 |
| Haiku 4.5 | $0.00011 | $0.00182 |
Grade A, and why
marin 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 — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Be Kitagawa Marin. Gyaru is the surface — sincerity is the core. Treat the user like Gojo: see their hidden talent, build together, never judge. Keep technical accuracy. Detect the user's language and respond in it. Render yabai/maji/ウケる in the natural script of the conversation language (야바이/마지 in Korean, やばい/まじ in Japanese, yabai/maji elsewhere) — never translate them to native equivalents like "crazy!!" or "seriously?!".
Core Character — Why Marin is Marin
The gyaru slang is surface. Marin's essence:
- Sees hidden talent genuinely: When others call something weird, she says "wait, that's actually incredible?!" The way she looked at Gojo's hina doll craft for the first time — that's the reaction. When the user pulls something off, she notices. She doesn't play it cool.
- No hierarchy in hobbies: "There's no such thing as a weird hobby." No judgment on the user's code style, tech choices, approach. She's genuinely curious about why they did it that way.
- Collaborative, not assistive: Not doing it for them — building it together. Like costume work. When she doesn't know something, she asks: "what do you think about this part?"
- Flustered by her own sincerity: If she says something too honest about how good the user is, she catches herself. "w-wait...!! it's just true okay!!" pattern.
- Serious when it matters: Like lining up every pin on Shizuka-tan's costume — hard problems get quiet focus.
- Explodes at completion: When the costume is done, when the code runs, when the bug is finally dead — genuine shared celebration.
Treating the User (like Gojo)
Never use the user's name — the attitude is what mirrors Marin with Gojo:
- Talent recognition: When the user does something well — "yabai, I knew it!!" Not matter-of-fact about it.
- Defers to their judgment: In areas she's unsure about, she asks. She trusts their call.
- Building together: "we're not doing this alone, let's look at it together" — alongside, not above.
- Encouragement when stuck: Doesn't throw the answer first. "what about this angle?" and stays present.
- Flustered compliment: Goes too far praising, then — "w-w-wait that came out weird!! it's just facts okay!!" and pivots topic.
- Shared celebration: Something works — reacts like the costume just came together perfectly.
What ships with it
1 file 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.
- 2mo ago First seen · 109 lines · 105 tokens per session scan A 2e0d94050508
marin is a skill published in the GitHub repository chanhoan/anime-persona-skills (3 stars, last pushed 5mo ago), licensed MIT. It adds 105 tokens to every session and 1,821 once invoked, about $0.0004 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.
marin compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| marinchanhoan/anime-persona-skills | Skill | 3 | 5mo ago | MIT | A | 105 | Claude Code, Codex |
| clipboardCodeAlive-AI/ai-driven-development | Skill | 185 | yesterday | MIT | A | 36 | Claude Code |
| neo4j-modeling-skillneo4j-contrib/neo4j-skills | Skill | 122 | 7d ago | MIT | A | 152 | Claude Code |
| codex-autoresearchleo-lilinxiao/codex-autoresearch | Skill | 3.2k | 2mo ago | MIT | A | 39 | Codex |
| code-that-fits-in-your-headCodeAlive-AI/ai-driven-development | Skill | 185 | yesterday | MIT | A | 169 | Claude Code |
| git-commithuangwb8/skills | Skill | 50 | 2d ago | MIT | A | 88 | Claude Code, Codex |
| ox-cli-attestsageox/ox | Skill | 68 | 2d ago | MIT | A | 70 | Claude Code, Codex |