SkillPedia Skills Encyclopedia
中文 EN
Today's updates 2026-11-18 UTC

No new commits yet today

Open the update feed

karpathy-skillalchaincyf/karpathy-skill

Breaks "evaluating an AI model or technical approach" into a checklist of questions that can be asked one by one,…

No ratings yet Rate this skill
Stars313
Forks84
Watchers313
Open Issues2

InstallOne-line install

Installable · 39 skills

This repository follows the Agent Skills spec, so one command installs it into your AI client — works across Claude Code, Codex, Cursor, Copilot, Windsurf, and 18+ other clients.

npx skills@latest add alchaincyf/karpathy-skill

39 more are not listed — run npx skills@latest add alchaincyf/karpathy-skill --list to see them all.

Add -g install globally (available in every project) · add --agent claude-code to target a client · add --yes to skip the prompts. Restart your AI client afterward and it will pick it up.

OverviewOverview

Key Features

Breaks "evaluating an AI model or technical approach" into a checklist of questions that can be asked one by one, so conclusions are not drawn from a demo alone. Key capabilities include: 1. Architecture Details: probes the model's architecture, training data, parameter count and compute cost, and points to technical reports and papers for verification; 2. Benchmark Performance: checks results on standard benchmarks and comparison against SOTA, requiring the latest evaluation results rather than marketing claims; 3. Code / Implementation: confirms whether an open-source implementation exists, its code quality and reproducibility, and whether it can be used directly; 4. Scaling Characteristics: analyzes how capability changes with scale and where the returns taper off; 5. Demo vs Deployment: separates demo performance from behavior under real deployment conditions; 6. March of Nines: reviews failure probability level by level using a "nines of reliability" lens; 7. Data Flywheel: assesses whether data feedback and iteration form a continuous improvement loop; 8. Competitive Landscape: maps comparable approaches and competitors to judge relative advantage; 9. Basic Facts: anchors the model's basic factual information first, avoiding analysis built on wrong premises.

Use Cases

Structured due diligence on candidate models before selection, writing model evaluation reports, deciding whether an open-source project is worth adopting, and explaining a model's real capability boundaries to a team or investors.

Target Roles

AI / Algorithm Engineer, Data Analyst / Data Engineer, Product Manager (PM), Full-stack / Indie Developer.

RatingUser ratings

Nobody has rated this skill yet. Yours would be the first on the site — your vote becomes the initial average.

Rate this skill

Fill it in to be included in the per-role averages below.

One vote per IP,and it cannot be changed after submitting, so choose carefully the first time. No account is collected and no sign-in is required.

RepositoryRepository

Skill ID2344
Data sourceGitHub
Repositoryalchaincyf/karpathy-skill
CategoryAI & Agent Skills / AI Technology Research & Evaluation
Primary languageOther
LicenseMIT License
Default branchmaster
GitHub repository ID1203030411
Node IDR_kgDOR7TJiw
Rank#1820 (position within the source ranking)
Site-wide popularity rank#2045 (computed live from Stars, across all 5833 entries)

GrowthStar / Fork growth

One snapshot per day, tracking how Stars and Forks changed

01002003004002026-09-252026-10-132026-09-25 · 3052026-10-13 · 313
  • Now313
  • Previous point · 09-25 305
  • vs previous point +8
  • 2 data points 18 days
02550751002026-09-252026-10-132026-09-25 · 842026-10-13 · 84
  • Now84
  • Previous point · 09-25 84
  • vs previous point ±0
  • 2 data points 18 days

TimelineTimeline

Repository created2026-04-06 16:46:277 months ago
Data last updated2026-10-26 12:28:2923 days ago
Last commit2026-08-25 03:05:062 months ago

CommitLatest commit

Version [latest] committed by alchaincyf. Update Category: Documentation & Guidelines Update. Commit details: docs: 版本自检机制——30天静默自查一次,仅确认有新版本时任务后提示. This update improves usage instructions, deployment guides, and developer docs.

0b10e85 vlatest alchaincyf 2026-08-25