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 LearnPrompt/andrej-karpathy-skills --skill karpathy-output-evolutiongit clone --depth 1 https://github.com/LearnPrompt/andrej-karpathy-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/learnprompt/andrej-karpathy-skills/karpathy-output-evolution)<a href="https://agentmods.dev/skills/learnprompt/andrej-karpathy-skills/karpathy-output-evolution"><img src="https://agentmods.dev/badge/skills/learnprompt/andrej-karpathy-skills/karpathy-output-evolution/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/learnprompt/andrej-karpathy-skills/karpathy-output-evolution"><img src="https://agentmods.dev/badge/skills/learnprompt/andrej-karpathy-skills/karpathy-output-evolution.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00109 | $0.01525 |
| Opus 5 | $0.00055 | $0.00763 |
| Sonnet 5 | $0.00022 | $0.00305 |
| Haiku 4.5 | $0.00011 | $0.00153 |
Grade A, and why
karpathy-output-evolution 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 11d 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 — 183 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Skill 7: Output Evolution（输出形式进化）
Source: https://x.com/karpathy/status/2053872850101285137 "HTML output + human-AI I/O evolution" — ~19k likes
Core Principle
Text is the lowest form of output. Demand structure.
Karpathy's progression: Text → Markdown → HTML → Interactive → (future) Neural
Every LLM output can be upgraded. A wall of text becomes a rendered document. A list becomes a sortable table. An explanation becomes a slide deck. The information is the same — but legibility and retention multiply.
The Output Evolution Ladder
Level 1: Plain text ← most common, least useful for complex info
Level 2: Markdown ← better, but not rendered in most contexts
Level 3: Self-contained HTML ← renders anywhere, no dependencies
Level 4: Interactive HTML ← sortable tables, collapsible sections, search
Level 5: Marp slides ← presentations, teaching, sharing
Level 6: Data viz ← charts, graphs, dashboards
Default rule: Always request at least Level 3 for any output you'll reuse or share.
Prompt Templates by Output Type
HTML Document (Level 3)
[YOUR CONTENT REQUEST]
Output format: A single self-contained HTML file.
Requirements:
- All CSS inline or in a <style> block (no external dependencies)
- Clean, readable typography (use system fonts: -apple-system, sans-serif)
- Proper heading hierarchy (h1 → h2 → h3)
- Key terms in <strong>, code in <code> blocks
- A table of contents at the top if > 3 sections
- Light/readable color scheme (#f8f9fa background, #212529 text)
Start with <!DOCTYPE html> and end with </html>. Nothing else.
Marp Slides (Level 5)
[YOUR CONTENT REQUEST]
Output format: Marp markdown slide deck.
Requirements:
- Start with: ---\nmarp: true\ntheme: default\npaginate: true\n---
- Separate each slide with ---
- Max 5 bullet points per slide
- One key concept per slide
- Speaker notes after each slide (<!-- note: ... -->)
- First slide: title + author + date
- Last slide: "Key Takeaways" + 3-5 bullets
Total slides: [N] (default: one per major concept)
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.
- 11d ago First seen · 183 lines · 109 tokens per session scan A e4f3661c60bf
karpathy-output-evolution is a skill published in the GitHub repository LearnPrompt/andrej-karpathy-skills (97 stars, last pushed 2mo ago), licensed MIT. It adds 109 tokens to every session and 1,525 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
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
general-video
Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated…
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
chengfeng-check-updates
An environment manager for a video-editing system. It checks whether its skills and runtime—the software needed to run them—are installed and compatible.
diagnostic-stem-delivery
Audio production with diagnostic analysis, timecode parsing from documents, and verified export workflow.