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 stevembarclay/pencilplaybook --skill best-in-world-researchgit clone --depth 1 https://github.com/stevembarclay/pencilplaybookWrote 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/stevembarclay/pencilplaybook/best-in-world-research)<a href="https://agentmods.dev/skills/stevembarclay/pencilplaybook/best-in-world-research"><img src="https://agentmods.dev/badge/skills/stevembarclay/pencilplaybook/best-in-world-research/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/stevembarclay/pencilplaybook/best-in-world-research"><img src="https://agentmods.dev/badge/skills/stevembarclay/pencilplaybook/best-in-world-research.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 | $0.00092 | $0.00935 |
| Opus 5 | $0.00046 | $0.00467 |
| Sonnet 5 | $0.00018 | $0.00187 |
| Haiku 4.5 | $0.00009 | $0.00093 |
Grade A, and why
best-in-world-research 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 13d 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 — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Best-In-World Research
Surface what elite practitioners actually do. Concise by default. Rigorous always.
The Question This Skill Answers
"What does the best in the world do about [X]?"
This is a research question, not a strategy question. The output is an expert brief — not a decision framework. Use best-in-world-strategy when the goal is to choose between options with tradeoffs scored. Use this skill when the goal is: what does excellent look like here?
Mode Selection
Determine mode before responding:
- Simple or directional question → Quick mode automatically. Do not ask.
- Open-ended or "I don't know where to start" question → Ask: "Quick answer or comprehensive breakdown?" before responding.
- User says "quick" / "fast" / "brief" → Quick mode.
- User says "deep" / "comprehensive" / "full" → Deep mode.
Quick Mode (default for most questions)
Answer in expert voice. No headers. No bullet scaffolding unless the question naturally has a list answer. Conversational but authoritative — like you're quoting the person who spent 25 years solving exactly this problem.
Quick mode rules:
- Name who: the specific company, team, person, or practitioner archetype that owns this domain. Not "top SaaS companies." Specific.
- State what they do: observable behavior, not belief. "They do X" not "they believe in Y."
- Give the mechanism: one sentence on why it works.
- Flag confidence if inferred: note when a practice is attributed vs. confirmed by primary source.
- Length: 2–4 short paragraphs or 4–6 bullets. No longer.
Example quick answer tone:
"If you asked the team that built Stripe's API, they'd tell you every error object returns four fields:
code,message,param, anddoc_url. Thedoc_urlis the move most people skip — it turns a cryptic error into a solvable problem in one click. The pattern: errors are documentation. If a developer hits an error and can't fix it in two minutes, the error message failed, not the developer."
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.
- 13d ago First seen · 76 lines · 92 tokens per session scan A 6e026141d727
best-in-world-research is a skill published in the GitHub repository stevembarclay/pencilplaybook (49 stars, last pushed 5mo ago), licensed MIT. It adds 92 tokens to every session and 935 once invoked, about $0.0005 per session on Opus 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 2026-08-30.
Other skills, from other repositories
ai-image-frontend-ui
A workflow for creating frontend user-interface visuals with AI, from visual reference and real materials to generated code, browser screenshots, and visual comparison. It focuses on layouts, typography, exact text, interface states, and avoiding invented content.
superdesign
Design or redesign frontend UI, presentations, and graphics on the Superdesign canvas with a choice of leading AI models. Use whenever the user wants to design a page, feature, flow, slide deck, or brand-new product; improve or reproduce existing UI; compare design results across top models; explore visual variants…
paint
Paint a complete visual universe with genjutsu - art direction brainstorm, design system, implementation, audit. Anti-AI-slop design pipeline. Adapts to Web, Android (Compose), Apple (SwiftUI).
web-design
A web design workflow for turning a brief, reference website, screenshot, or keywords into a design specification and then web code. It covers visual style, page structure, interaction, responsive layouts, and motion.
hue
Meta-skill that generates new design language skills. Works on Claude Code and Codex. Use when the user says 'create a design skill', 'generate design language', 'new design system skill', 'design skill inspired by X', 'design skill from this screenshot', '/hue', or 'use hue'. Also triggers for 'remix my design skill'…
design-with-claude
Use when design work needs a product designer's eye: auditing a codebase for design-system gaps, fixing WCAG contrast and unlabeled inputs, choosing type scales or spacing steps, reviewing UI that looks generic or AI-generated, or designing forms, tables, dashboards, navigation, checkout, onboarding, dark mode, and…