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 big-emotion/agent-atelier --skill expertifygit clone --depth 1 https://github.com/big-emotion/agent-atelierWrote 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/big-emotion/agent-atelier/expertify)<a href="https://agentmods.dev/skills/big-emotion/agent-atelier/expertify"><img src="https://agentmods.dev/badge/skills/big-emotion/agent-atelier/expertify/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/big-emotion/agent-atelier/expertify"><img src="https://agentmods.dev/badge/skills/big-emotion/agent-atelier/expertify.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 | $0.00095 | $0.00919 |
| Opus 5.5 | $0.00038 | $0.00368 |
| Sonnet 5.5 | $0.00019 | $0.00184 |
| Haiku 4.5 | $0.00010 | $0.00092 |
Grade A, and why
expertify 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 — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Expertify — proper role + domain terminology
Turn a layperson's prompt into an expert-grade prompt: find the professional role that owns the subject, swap lay wording for the field's established terminology, and produce a rewritten prompt that opens with a persona line.
Input
$ARGUMENTS is one of:
- a rough prompt the user intends to ask ("help me make my website load faster on phones"),
- a topic they want the vocabulary for ("retirement savings in France"),
- a fuzzy role description ("the person who decides which features get built").
If $ARGUMENTS is empty, ask the user for the prompt or topic to expertify, then proceed.
Workflow
1. Identify the domain and the role
- Extract the subject domain from the input. If the input describes a person/function rather than a subject, map the description to the actual job title.
- Pick one primary role — the most precise title an expert in this exact question would hold. Prefer the specific over the generic: "ophthalmologist" over "eye doctor", "SRE" over "sysadmin", "notaire" over "legal person", "conversion rate optimizer" over "marketing person".
- If two roles genuinely split the subject, keep one as primary and mention the alternative in one line.
2. Verify when unsure
- Mainstream domain and you are confident → use your own knowledge, no search.
- Niche domain, ambiguous role, possibly evolved terminology, or country-specific titles (FR vs EN job markets differ) → run 1–2 web searches (e.g.
<domain> job titles,<domain> glossary terminology) and cross-check before answering. Do not over-search; two queries max unless results conflict.
3. Build the terminology map
- List the lay terms in the user's input and pair each with the field's proper term.
- Add the key concepts an expert would name that the user did not know to ask about — these are often the most valuable entries.
- 6–12 entries. Every entry must be relevant to the actual input; no glossary padding.
- Never invent jargon. If no established term exists for something, say so explicitly.
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 · 60 lines · 95 tokens per session scan A 1b75b7b43d38
expertify is a skill published in the GitHub repository big-emotion/agent-atelier (3 stars, last pushed 23d ago), licensed MIT. It adds 95 tokens to every session and 919 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.
expertify compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| expertifybig-emotion/agent-atelier | Skill | 3 | 23d ago | MIT | A | 95 | Claude Code |
| skill-meta-promptnyldn/claude-octopus | Skill | 4.4k | yesterday | MIT | A | 19 | Claude Code |
| ai-workflow-architectcbrock84/headcount | Skill | 2.6k | 19d ago | MIT | A | 89 | Claude Code |
| cloudflare-workers-aisecondsky/claude-skills | Skill | 245 | 6d ago | MIT | A | 39 | Claude Code |
| nano-banana-promptingNikiforovAll/claude-code-rules | Skill | 142 | 4d ago | Apache-2.0 | A | 51 | Claude Code |
| nano-bananaNikiforovAll/claude-code-rules | Skill | 142 | 4d ago | Apache-2.0 | A | 73 | Claude Code |
| cortex-prompttonone-ai/tonone | Skill | 77 | 23d ago | MIT | A | 55 | Claude Code |