Developer Tools
andrej-karpathy-skills
Last updated 2026-06-05 · benchmark checked 2026-08-10 · unchanged since 2026-08-08 — deterministic & reproducible
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
Rated 4.17★ by 6 on Legit.Show.
Is andrej-karpathy-skills production-ready?
Legit.Show scores andrej-karpathy-skills 38 out of 100 — the simple average of its 3 measured frames. Legit.Show ran its deterministic 7-Frame production-readiness benchmark on andrej-karpathy-skills (github assessment), measured from the public surface with no LLM in the scoring path. Its strongest frame is Security; its weakest is Discoverability. 3 of the seven frames returned a score; Performance, Accessibility, Privacy and Reliability were not measurable on this service and are recorded as null — not as zero. Maintenance is an additional frame from the open-source teardown, scored separately from the seven. Every frame it averages is published with its evidence on the Legit.Show listing.
The 7 Frames
- Performance — not measurable on this service (null — not scored as 0)
- Accessibility — not measurable on this service (null — not scored as 0)
- Security — 40/100
- Privacy — not measurable on this service (null — not scored as 0)
- Reliability — not measurable on this service (null — not scored as 0)
- Standards — 40/100
- Discoverability — 35/100
Open-source teardown
Scored separately — not one of the seven.
- Maintenance — 55/100
What we measured
- No Content-Security-Policy and no HSTS.
- No privacy policy found.
- Sets cookies / loads scripts with no consent prompt.
Who built it
Andrej Karpathy
Visit andrej-karpathy-skills → · Alternatives to andrej-karpathy-skills → · How this was measured →